RELX PLC Sponsored ADR Stock

Great. Thank you, everybody. Good afternoon and for some, good morning. Welcome. I'm Paul Sullivan, Head of Investor Relations at RELX, and it's a pleasure to welcome you to our STM Investor Seminar today and also introduce the CEO of Elsevier, Kumsal Bayazit. Kumsal?

Thank you, Paul. Good afternoon, and good morning to those joining us from the U.S. Welcome to the STM Investor Seminar. Thank you very much for joining us today. I'm Kumsal Bayazit, CEO of Elsevier, the STM division of RELX. I joined RELX in 2004 and have been the CEO of Elsevier for the last 8 years. Before that, I held operational and strategic roles in all parts of RELX in risk, in legal, in exhibitions as well as at RELX Corporate as Chief Strategy Officer and Chair of RELX Technology Forum.

It has been 4 years since we last did a deep dive on Elsevier. My team and I are really looking forward to sharing with you the progress we have made since then as well as the significant opportunity that we see ahead.

I will start with an overview of the business and our strategy. Gemma Hersh will then cover Primary Research. I will come back to cover our decision support tools and 2 of our product leaders, Max Khan and Paul Crockett, will bring these to life with product demos. We will then hear from our Chief Technology Officer, Jill Luber, and I will close with some concluding remarks and move to Q&A.

Let me start by putting Elsevier in context. In H1 2026, STM represented 28% of RELX revenue and 30% of adjusted operating profit. That equates to annualized revenue of roughly $4 billion or GBP 3 billion and adjusted operating profit of over GBP 1 billion. We are a very global business with customers in more than 170 countries and territories. A little over 40% of our revenue comes from North America, a little over 20% from Europe and just over 1/3 from the Rest of the World, which gives us broad exposure to research and health care spend around the globe.

Almost 80% of revenue is subscription-based, primarily under multiyear contracts. And the rest is transactional, much of it recurring in nature. The large subscription base of the business gives us stability. It also means the positive impact of increasing volumes or new higher-value products comes through gradually in our results. A little over half of STM's revenue is in Primary Research. Here, we offer article publishing, research and analytics platforms. Decision support tools, which we also refer to as databases, tools and electronic reference, is around 40% of revenues. What we do here is help our customers with evidence-based decisions and tasks through multiple offerings.

The 2 parts of the business reinforce each other. Primary Research gives us the content, the credibility, the domain expertise and the expert community on which our decision support tools are built. And our decision support tools increase value for our customers by combining their trusted content and curated data with advanced analytics and evolving technologies such as artificial intelligence.

We operate in large global segments with attractive structural long-term growth drivers. R&D spend is critical for nations and corporations to drive innovation, economic growth and solve societal issues such as finding cures for diseases or climate change. This is why in real terms, global R&D spend has consistently grown at an annualized rate of 5% over decades. In nominal terms, the growth is, of course, higher. This growth trend has been secular and has held through economic cycles, geopolitical tensions, political and policy changes.

In any given year, of course, there can be ups and downs in spending in any one country or institution or discipline given budget pressures or short-term fluctuations. But fundamentally, over the long-term, global R&D spend continues to grow. This is one of the reasons the number of researchers also consistently grows at about 5% per year. Now I could have shown you this data for 50 years, and you would have seen the same trend. And we expect this growth to continue for many decades to come.

Technology advancements have always improved research productivity. It's another driver of growth. And generative AI is having a similar impact. These drivers, R&D spend, number of researchers and improving researcher productivity lead to more research, more discoveries, more information and content, higher number of submitted and published articles and greater demand for more evidence-based decision support tools to understand, build and apply on knowledge.

Looking at health, you can see very similar trends with global health expenditure growing at 3% a year plus inflation. The number of physicians and nurses has also been consistently growing globally. We serve a large global customer base. We work with over 10,000 academic and government institutions, including virtually all leading research organizations worldwide. We do extensive work with millions of researchers who interact with our tools each year. Our teams have deep connections with research leaders and leading researchers in their fields.

We also serve over 2,000 R&D-intensive corporations and work with leading companies in life sciences and biopharma, engineering and technology, chemistry and energy. In health, we have solutions for most stakeholders in the health ecosystem from medical and nursing schools to health care providers such as hospitals. No single customer accounts for more than roughly 1% of our revenue. And in every segment, our objective is to be a trusted partner and to help customers improve their outcomes so they can, in turn, advance science and health care.

Our relationships are long-term and deeply embedded in how our customers work, which is a strong foundation for our decision support tools. We have 3 main strategic priorities. First, we expand customer value in Primary Research. We continue to deliver industry-leading research quality, article volume and user experience on leading technology platforms at a lower effective price. Technology advancements such as AI enables significant improvements in the author, editor and peer review experience and outcomes throughout the whole editorial process.

Second, our decision support tools continue to deliver higher customer value. By combining our deep domain expertise, leading evidence-based scientific content, advanced analytics and evolving technology, we deliver innovative products that improve customer outcomes in existing and new use cases. And you will see examples of that with LeapSpace and Sherpath today.

Third, our global modular data and technology platform enables us to build segment-leading products more efficiently. We have always operated with a technology-agnostic approach so that we choose the right tools for each use case as technology evolves. A current example of that is our multi-model AI orchestration layer, so we can optimize for quality, latency and cost. We also leverage our technology platform internally to accelerate our speed and agility and to drive efficiencies. That is what allows us to innovate at pace and manage cost growth below revenue growth. And these priorities reinforce each other.

Now at our last seminar, I spoke about how we evolved the business towards higher-value, high-growth decision support tools, which had improved our revenue growth from 2% to 4%. Since then, we have continued to deliver on that improving long-term trajectory, and we are currently at 6% growth. Equally importantly, during this time period, our customer metrics, including Net Promoter Scores for products and buyers and customer satisfaction have been on an improving trajectory. Employee engagement metrics have also improved.

Now let me talk you through the drivers of our improving revenue growth in a bit more detail. 4 years ago, we said that Primary Research had been growing in the low single digits. Since then, we have delivered increasing growth in submissions, high-quality published articles with an expanding journal portfolio and improved author, editor and peer reviewer experience. That volume growth has started to come through gradually in revenue growth, and we are now growing in mid-single digits in Primary Research.

For decision support tools at our last seminar, we said revenues were growing in mid- to high single digits. As we add more customer value with increasing pace of product innovation, we are now delivering strong growth in this segment. Our objective is to continue on our improving long-term growth trajectory by delivering gradual improvements in both Primary Research and decision support over the long-term. Our objective is also to continue to manage profit growth ahead of revenue growth through process innovation. So I hope that this gives you a good overview of STM.

I'm now going to hand over to Gemma for a deep dive on Primary Research.

Thank you, Kumsal, and hello, everybody. I'm Gemma Hersh, Senior Vice President, Primary Research at Elsevier. I've been with Elsevier for over a decade, and I've held a variety of strategic and commercial roles. Today, I'll cover what is it that we do in Primary Research, the value we create for the authors, reviewers, editors and readers we serve, the opportunity in front of us and what we're doing to capture it.

At the heart of what we do in Primary Research is supporting researchers and their discoveries. Researchers get funding based on their research proposals from governments, funders and corporations. They conduct the research at the end of which they have findings. They codify these findings in an article and look for the right journal in which to share them. Publishing plays a critical role in how researchers share their contribution with the scientific community and build their reputation and careers.

Published articles also enable other researchers to build on this knowledge. We enable that process by helping researchers verify and share their findings. The process starts with the researchers submitting their article to the journal they'd like to be published in. All submitted papers are first assessed by one of the editors of the journal. These editors are recognized experts in the field. Some work as full-time employees, others work part time alongside roles within academic or research institutions.

In total, we work with over 37,000 editors, supported by around 105,000 editorial board members. We have publishing teams who constantly recruit leading experts in their fields as editors and editorial board members. For example, over the past 12 months, we recruited approximately 8,000 new editors and 12,000 editorial board members. Around 70% of submitted articles do not progress beyond that first assessment called the Editorial Review, either because they're not the right fit for the journal or because the science is not yet sound.

Papers that progress go to the second assessment called Peer Review. This means editors send the article to at least 2 reviewers who are experts in the specific field. In total, we work with more than 1.9 million reviewers per year. During the peer review process, the author will respond to questions, improve the paper and sometimes conduct additional work. Peer review pressure tests research findings. Expert peers scrutinize the methods, challenge the conclusions, their feedback improves papers, sharpens arguments and highlights gaps so the published record is stronger.

At the end of peer review, about half of the manuscripts they received are accepted for publishing, meaning less than 20% of all articles that are submitted to Elsevier are published. This process provides quality control so that other scientists can trust the findings in published papers. Clinical guidelines, drug development and policies depend on peer-reviewed published findings. The published article becomes the version of record that clinicians, funders and policymakers can rely on and other researchers can build on.

After the article is accepted by the editor and before it's published, it goes through the production stage. It's critical that the paper does not have any mistakes and all images and data are accurate. During this stage, we also add metadata that enables the extraction of billions of data points. These data points also fuel the quality and value of our decision support tools for customers. Then we distribute the article via our research platforms. ScienceDirect is the biggest with significant reach, 20 million unique visitors monthly, which drives broad visibility and readership.

This is critical for researchers as it enables other researchers to build on their findings, helps drive collaboration, funding and ultimately supports researchers' career goals and their impact on society. We spoke about this in the last seminar. It's easy to launch a journal. It is much harder to build a mature one with a trusted community at scale and with global reach. On average, it takes 4 years for a new journal to get indexed by a trusted indexation service, and it usually takes much longer than that to become a leading journal in a field.

Kumsal touched on how technology advancements have had a positive impact on research productivity. AI and generative AI are great examples of this. They speed up the research process and make it more efficient. That means more quality findings more quickly and more submissions to publish. However, it also means more research integrity issues and AI slop, which makes trusted journal brands and trusted peer review processes even more valuable.

Journal quality is a key driver of submissions and citations are a proxy of that. Journals that are the most highly cited develop reputations for being impactful and high quality over time. We have the largest and the highest quality portfolio and scale and quality both matter. For authors, publishing in a high-quality journal raises the visibility of their work. We look at quality in 3 ways: first, share. We publish 18% of all articles and receive around 30% of citations, which means the research we publish is read and cited more frequently.

Second, field-weighted citation impact, which normalizes citations for differences between scientific fields. The industry average is 1, Elsevier is 1.46, well ahead of other large publishers, as you can see on this chart. Authors, funders and institutions use measures like these when deciding where to publish, what research to fund and to measure the impact of that research. The third lens is quality tiers, where journals are ranked by field-weighted citation impact.

We participate in every tier because our objective is to find a home for all sound science, but our portfolio is weighted to the top. We publish about 40% of all articles in the top 10% of journals by this measure. This was 32% when we presented to you in 2022. So we've grown our article share in this top quality tier. For our authors, that means a wide choice of high-quality journals in their discipline. For customers, it means a subscription that gives access to a portfolio weighted to the highest impact research. This is a virtuous cycle.

A high-quality portfolio attracts high-quality submissions because authors want to publish in trusted journals in their discipline. And that content with our share of citations is also the foundation on which decision support tools are built. Quality is more important today than it has ever been with increasing research integrity issues and AI slop, and we expect it to continue to be even more important in the future.

Let me now turn to the opportunity and then to how we are capturing it. Today, total industry articles published has grown at a compound rate of 5% a year since 2015, and we expect that growth to continue. We covered the drivers of this growth, continued long-term growth in R&D spend, number of researchers and advancing technology such as AI, improving research productivity. Zooming in on Elsevier, our goal is to continue to grow faster than the industry.

Let me start with submissions. Elsevier submissions have grown from 1.3 million in 2015 to 4.2 million in 2025, a compound annual growth rate of 12%. Growth was over 20% in 2025 and over 20% again in the first half of 2026. So momentum remains very strong. We expect continued strong submission growth for many years to come. Why are authors choosing Elsevier? Well, first, we have a very extensive portfolio with over 3,000 journals and industry-leading quality. Journals like The Lancet have been a leader in global health for over 2 centuries. Journals like Cell in biology and Blood in medicine are field shaping. Concerns about research integrity may also be leading more authors to choose established, trusted brands with a long track record.

We're also continuously improving user experience for authors, editors and reviewers. We give authors choice in how they publish so they can publish in a journal that meets national or institutional policies that they need to follow. Published articles have grown from 400,000 in 2015 to over 800,000 in 2025, a compound annual growth rate of 7%. Over any multiyear period, our published articles grow more slowly than submissions by design. We are selective with a rigorous quality-focused selection process across all our journals, and that is important to protect research integrity and our quality position.

We have 3 ways of growing published articles while protecting quality. First, we're actively extending our portfolio across all disciplines and tiers. We launch approximately 70 to 80 new journals a year to fill gaps in disciplines or quality tiers or in emerging fields. One example of this is Chem Circularity, one of our new Cell Press journals, which advances closed-loop solutions across chemistry, materials and engineering. In addition to launching new journals, we also focus on successfully growing and maturing our existing journals.

Second, we support authors in trying to find a home for all articles that are submitted to one of our journals, even if they do not get accepted in their first journal of choice but are suitable to be published in another journal. We call this article transfer service. It takes a lot of effort to facilitate this at scale. We use technology and automated processes as well as increasing editorial capacity to help authors find a suitable journal to publish their article. It's important to highlight that it's always the author's choice on where to submit.

Third, we're constantly improving our submission and peer review systems to improve the author experience and develop tools to support reviewers and editors to make the best use of their time and enable them to make better decisions more efficiently. This lets us increase our volume without compromising quality or research integrity, and I'll come back to this in a minute.

Now to the value customers receive. Our strategy in Primary Research has been consistent, higher quality and a lower average effective unit price in the industry. The 2 primary payment models in Primary Research are pay-to-read and pay-to-publish. Both models are available as a subscription or on a transactional basis to our customers. We provide both payment models for our services as well as combinations of the 2 to support our customers' diverse needs and preferences.

As our article submissions have grown by double digits, our volume of published articles has grown by mid- to high single digits, and our Primary Research revenue has grown below this. Revenue per article, which is a proxy for average industry spend per article that we publish has decreased consistently for many years. As you can see on this chart, we have continuously delivered more customer value for every dollar spent while increasing our revenue growth for the last decade. This is our strategy, deliver more customer value at a lower effective unit price and maintain high levels of quality. This ensures customer trust and revenue growth.

Let me show you now how we add significantly more value to Primary Research by leveraging sophisticated analytics and advanced technology like AI. The first value is safeguarding research integrity and maintaining trust. Integrity issues in research range from honest mistakes to deliberate misconduct, AI slop to fraudulent manuscripts, including organized paper mills. We screen every article from before submission to after publication with many active detection signals covering misuse of AI, policy compliance, authorship manipulation, citation abuse, paper mill activity and data fabrication.

We continuously increase the number and quality of the signals, and we draw on our fraud expertise from our risk division, giving us richer integrity capabilities. The more papers we see, the more we can identify new signals of misconduct and so the better our research integrity solutions get. Our new tools bring screening, investigation and decisions into one system. So every case moves from identification to resolution consistently and every decision is transparent, traceable and defensible. We expect this dynamic of AI slop and research fraud to continue to be an arms race like fraud in other industries. We will continue to mature our tools and drive innovation to ensure that the scientific record can remain trustworthy.

The second area of value is to continue to evolve editorial and peer review processes to drive higher quality, faster turnaround and lower cost. We have already made good progress and are seeing measurable results. Let me give you 2 examples that are quite advanced in implementation. The first one is a vastly improved submission process, which we accelerated the rollout of about a year ago. The result is faster, better quality submissions. Median submission time by an author is down 60% and 90% of authors find the new experience both useful and easy to use. Editors benefit too because they receive better quality submissions. This has been rolled out to nearly 2/3 of our 3,000 journals.

The second example is a fundamentally redesigned peer review experience for which we also accelerated rollout this year. The result is more transparency, less friction and more positive peer review outcomes. User preference is around 85% among authors and reviewers. And the acceptance rate of peer-reviewed manuscripts has consistently improved. This also has been rolled out to more than 2/3 of our 3,000 journals. The implementation of these new tools is a significant change management process with tens of thousands of editors and millions of authors and peer reviewers.

Let me give you 3 examples that are in testing stage and that we plan to gradually roll out. For editors, we're launching AI-enabled tools to support them in their work so that they spend less time gathering information and more time making decisions. Over time, we see an opportunity to automate more of editorial triage so editors can focus on decisions that genuinely require human expertise.

Another example is from the acceptance to publication process where we prepare accepted papers for publication. Over the last decade, we cut the time for this process by more than 50%. We aim to decrease the time it takes by another 75% over the next 3 years. This will drive quicker access to the newest research, a more positive author experience and decreased costs. We are also building AI agents to do significantly more enrichment of the published content in this last mile. That drives richer analytics capabilities in our decision and task support tools as well as lower cost to serve for analytics use cases.

We're also in the process of rolling out LeapSpace capabilities such as Claim Radar, which Max will share shortly, to our editors, authors and peer reviewers. This is a very helpful tool during the editorial process to assess the paper. Across all these stages, human expertise remains central. Editors, reviewers and subject matter experts make the decisions and technology gives them better information, makes the assessment easier and quicker and frees up time to apply human judgment.

Now let me hand back to Kumsal.

Thank you very much, Gemma. Thank you. Okay. Well, let me sum up Primary Research. Long-term segment growth continues to be driven by volume growth with increases in number of researchers, R&D spend and continuing improvement in researcher productivity. We're growing faster than the industry because authors value our quality, reputation and reach.

Our strategy is to ensure that customer value increases continuously, high quality and trust, more research, faster time to publish and a better and more efficient editorial process at a lower effective price. We've already gradually improved revenue growth from low single digits to mid-single digits. Our objective in Primary Research is to continue to deliver gradual improvements in growth over the long-term.

Let me now cover our decision support tools. Now we provide a broad range of decision support tools to help our customers improve decision-making, increase productivity and achieve better outcomes. Collectively, they account for approximately 40% of our revenues. Research solutions are just under half of our decision support revenue, and it is a very global revenue base.

The use cases here focus on research tasks and management, and we serve 2 broad categories: tools that help institutions manage and evaluate research and research funding and tools that support scientific research, publishing and corporate research. Customers are research institutions in academic and government segments as well as R&D-intensive corporations. And within this category, roughly 1/3 of our revenue base comes from corporate customers.

Nursing solutions are around 1/4 of decision support revenue. These tools support the whole nursing journey from helping students build knowledge and prepare for licensure exams to supporting them as practitioners. This business is primarily in North America. The remaining quarter covers medical education, physician care, specialist use cases and patient engagement, split pretty evenly between the U.S. and the Rest of the World. Our objective is to increase customer value by improving existing use cases, expanding into new use cases and embedding where customers work.

This improving customer value translates to an improving revenue growth trajectory in the following 4 ways. The first is by increasing the value of our existing high-value tools. We're doing this by adding an AI layer to many of these products. Scopus AI is a good example of this. Revenue growth increases as more customers opt into these higher-value capabilities, and there's a spend uplift when they opt in usually in double digits.

The second is when we launch a completely new product. These offerings extend to new use cases as well as go deeper in existing use cases by drawing on the full strength of our capabilities. LeapSpace is a good example of this. These are new propositions that drive measurable value and improved outcomes for our customers. They also reinforce the value of our existing products and drive greater usage of both our partners and our content. These new products are priced by customer segment based on the value they provide. When our customers decide to buy these products, the additional new business improves our revenue growth.

Third is when we bring together different products and capabilities in our portfolio for a specific user segment and make these capabilities interoperable, easier to use and easier to integrate using AI. Our nursing education suite is a good example of this. Revenue growth improves as customers buy more products from us as they see additional value in the combined interoperable suite.

Fourth, we're also increasingly integrating where customers work. The combination of customers' data and our own content and data create higher value. For example, in corporate R&D, we integrate our products into our customers' agentic environments. These are machine-to-machine value propositions similar to our risk business and are priced based on value. As we deliver one or more of these value drivers for our customers, customers buy more of our products to capture that value and our revenue growth improves.

We provide higher customer value by leveraging 4 key capabilities. The first is deep customer understanding and domain expertise. There are several factors here, decades of close partnerships with our customers, deeply ingrained expert communities of editors, editorial board members and peer reviewers across discipline. Many of our thousands of employees are researchers, PhDs, doctors and nurses, so have in-depth understanding of our customers' use cases. And finally, human oversight is ingrained in everything we do from content curation and integrity checks to expert-led evaluation of machine outputs.

The second is leading content and data. As Gemma just shared, our research content is of the highest quality. And in addition to our own content, we collect, license and curate large content and data sets, including abstracts, full text articles, patents, grants, policies, preprints and books. We continually expand the breadth and depth of our content in science and health care.

Third is sophisticated linking and analytics. We have extracted billions of data points from leading scientific and medical content sets. We have dozens of taxonomies and ontologies that create structured knowledge graphs and a semantic layer supporting scientific reasoning and evidence. This has been built over decades with deep domain expertise, and it would be hard and costly to replicate. This third capability enables us to provide benefits to our customers with outputs that have more transparency, more provenance and more consistency. Given we've already extracted billions of data points and pre-computed responses for a lot of different use cases, it also increases speed of response for our products and reduces costs, including token costs.

The fourth is advanced technology and global platforms. Our architecture is flexible and modular. So we build once and reuse everywhere. We can ingest new content and data and plug in new AI models seamlessly. We're technology-agnostic and multi-model. We have used extractive AI for more than a decade and now apply generative AI across our solutions in a way that supports the determinism and the accuracy that science and health care requires.

Transparency is built into product design. So outputs are more reproducible, more comprehensible, more defensible and linked to verifiable sources. Our solutions are private and secure, and we do not use customer data to train models. You will see each of these capabilities come alive in today's product demonstrations. While each capability is critical on its own, it's the unique combination of the 4 that creates the customer value.

I'm going to hand it over to Max now to provide a demo of LeapSpace. Over to you, Max.

Thank you, Kumsal, and hello all. I'm Max, Senior Vice President responsible for our Research Solutions.

I joined RELX in 2007 and over the last 15 years, I've had the privilege of serving our customers at Elsevier. We support the entire research journey by helping funders and policymakers decide where to invest and track outcomes so they can show taxpayers the value of their investments. Academic leaders shape their strategy, attract funding, manage faculty recruitment and serve their communities. Researchers develop research questions, review the evidence, find collaborators, plan and conduct studies and write and publish their findings. And corporate researchers apply research and R&D to support their innovation. Our solutions help people we serve across the entire research ecosystem get to better outcomes faster and in ways they can trust.

As Kumsal described, our research solutions improve outcomes for existing tasks and support new ones. LeapSpace is one example focused initially on researcher tasks. It helps them complete a wide range of tasks in one workspace, including developing ideas, planning research and writing up findings. LeapSpace is Elsevier's AI-enabled research workspace. We launched it in January 2026, building on ScienceDirect AI, which we introduced in March 2025. We added significant new capabilities in June and September, and we'll continue to expand them rapidly.

LeapSpace helps researchers by improving outcomes for tasks such as exploring ideas, planning research, finding funding and collaborating, all in one workspace, drawing on comprehensive full text content from Elsevier and other publishers, Scopus data, funding data and content that users can bring in themselves. Letting researchers check the evidence behind outputs, judge how much confidence to place in them and decide what to investigate next. We will work closely with researchers to keep improving LeapSpace, adding capabilities and supporting more tasks and types of users over time.

I'm now going to demo LeapSpace. Let's follow a cancer researcher through 3 tasks: understanding existing research, comparing approaches for an experiment and strengthening a paper before publication. I'll explain what's happening on the screen for each of the 3 tasks and how LeapSpace helps the researcher get to a better outcome on each task.

In the first task, the researcher wants to understand a treatment pattern for a particular cancer. They enter that question into LeapSpace and can also add their own materials. Here, the response brings together evidence from full text peer-reviewed content from Elsevier and other publishers as well as data from our Scopus database. LeapSpace is private and secure, and we do not use customer data to train models. Customers have told us that they value LeapSpace for bringing comprehensive peer-reviewed content and responsible AI together in one place while supporting research integrity, transparency and confidentiality.

Now that LeapSpace has provided a response, the researcher can check the evidence behind a particular finding. This panel, which we call a Trust Card, shows the study, who carried it out, how it relates to the finding and the exact passage behind the finding from a 10-page research paper. Researchers tell us that Trust Cards help them judge how much confidence to place in the finding, where to dig further and how to use it. As they consider the response, a researcher might want to check whether the wider research supports a particular claim before building on it.

This view brings together studies that support a claim, show mixed findings or contradict it along with a summary explanation. We call this Claim Radar. Gathering these studies can be time consuming. Here, the researcher can go straight into examining the evidence. Researchers tell us that they value being able to check references and understand conflicting findings. This supports their critical thinking and scientific judgment and keeps them in control of how they use the information.

Both Trust Cards and Claim Radar draw on scientific content and data we have already organized and linked to knowledge graph, connecting scientific concepts and evidence. This helps researchers trace findings to their sources while reusing the work already done helps us respond faster, more accurately with lower and more predictable processing costs.

In the next task, the researcher wants to plan an experiment, for example, to test an approach that targets cancer cells. The response brings together approaches used in previous studies, so the researcher can begin comparing how those experiments were carried out. Here, the researcher can compare studies side by side, including their methods and test conditions. This helps them spot differences that may explain different results and decide which approach and conditions to test next or how to best set up their own experiments. They can also add study details for closer comparison. Researchers tell us that comparison work, which used to take days or weeks can now be done in minutes.

In the final task, the researcher is writing a paper. They need to explain how their finding fits with existing research and make sure their argument is well supported by evidence. First, LeapSpace checks the researcher's argument against literature. LeapSpace then shows where each claim in the argument is supported and where evidence might be missing or needs further consideration. This helps the researcher decide which claims might need more support before publication. LeapSpace can then help strengthen the argument with further evidence. The researcher chooses to accept or reject every change. LeapSpace never silently rewrites the manuscript and the researcher stays responsible for their work.

A researcher once told us this writing support helped her strengthen her research proposal, surfacing papers she had missed and highlighting arguments that needed more support. Her proposal has been approved, and she's now writing a detailed plan for carrying out her study. LeapSpace integrates with Microsoft Word, so researchers can check claims using LeapSpace while they write. This is how we bring our tools into the places where customers work. Researchers must still follow their journals and institutions' AI policies.

As Gemma mentioned, we're also in the process of rolling out Claim Radar and other LeapSpace capabilities to editors, authors and peer reviewers. This helps them examine the evidence behind the paper's claims and identifying points that need closer review. You've now seen the 4 capabilities that Kumsal described working together. Deep customer understanding and domain expertise shape the tasks we focus on, leading content and data provide the evidence, sophisticated linking and analytics make that evidence traceable, transparent and explainable and advanced technology in global platforms let us choose the right AI models for each task and bring these agentic capabilities into one workspace and into tools such as Word. Together, they help researchers get to better outcomes faster while keeping scientific judgment with the researcher and the researcher in the driving seat. We'll continue to work closely with researchers to improve how LeapSpace supports their work.

We're seeing very positive feedback from customers on the value we bring in 3 primary areas. First, better outcomes on important research tasks. Researchers tell us LeapSpace helps them strengthen their work. Almost 90% reported help in formulating or refining research questions and hypotheses.

Second, greater efficiency. In a survey of users, 97% reported time savings and more than half reported time savings greater than 50%. Researchers described reinvesting that time in other tasks such as ideation, deeper reading and actually doing the research.

Third, trust and control. There's a trust gap for researchers using AI tools. Our industry surveys show that more than 80% of researchers use generative AI, but only 22% trust it. You've seen how LeapSpace bridges the trust gap by allowing researchers to inspect the sources behind the answers, examine the supporting evidence and calibrate their confidence. More than 85% of researchers using LeapSpace tell us that the answers are supported by robust evidence.

The way that customer value drives our revenue for LeapSpace is that we sell it as a stand-alone product, primarily a new subscription to institutions. In academic and government, pricing is based on research intensity, the number of researchers and geography. For corporate customers, the pricing is value-based. LeapSpace was only launched a few months ago, so it's still small compared to our overall revenue.

Academic institutions typically have long sales cycles that are usually tied to 12-month academic calendars when budgets are set for the upcoming year. And as you know, LeapSpace hasn't been in the market for a full calendar year yet. Corporate customers typically have shorter sales cycles and faster decision-making processes, and we're seeing faster take-up. Across these customer segments, we already have roughly 1,000 institutional customers who have subscribed and monthly active users and usage have continued to grow rapidly in these accounts. In addition to our paid users, we've been focusing on building awareness.

We are fortunate to have millions of touch points with researchers via the editorial process and our destination sites such as ScienceDirect. Since June, we've been surfacing LeapSpace capabilities and features to showcase the value in the right moments for researchers. Through these activities, over 1 million users have already engaged with one or more features on LeapSpace. That helps build awareness and extends the feedback on the users and our product experience. So despite the early days, we're very excited about the future opportunity.

Now over to Paul for the next demo.

Thanks, Max. I'm Paul Crockett, and I'm the Senior Vice President for our Healthcare Education business. I've been working in education, publishing and technology for over 30 years. And prior to Elsevier, I was the CEO and Founder of Authess, an EdTech start-up acquired by Elsevier in 2020.

Our nursing solutions serve educators, nursing students and nurses throughout their careers, providing solutions to support nurses as they build foundational knowledge, prepare for licensure exams and finally, supporting them when they move to practice, applying clinical skills at the point of care. Our nursing business is primarily in the U.S., and there are close to 5 million nurses in the United States. With increasing demands on health care and a projected 8% shortage of registered nurses by 2028, nursing is a critical education pipeline needed to support labor demand. Today, I will cover our nursing education portfolio, which is a bit over half of our nursing solutions revenue.

So we provide a suite of connected solutions that help drive better outcomes and greater efficiency in nursing schools. Foundational to our solutions is our content that is built and continually updated by experienced nurse practitioners and educators. We utilize over 15,000 lessons, 500,000 adaptive quiz questions, 100,000 licensure assessment questions and more, all built upon our decades-long brand and content strength of nursing education books. These assets ground our AI-enabled teaching solutions, which help students build knowledge, apply clinical judgment and assess and review their learnings as they prepare for exams.

We also offer a suite of predictive analytics that help faculty understand the student's progress and their likelihood of passing the licensure exam. These are based on proprietary data sets built over many years, and our analytics give teachers an early view on where and when to intervene. It also creates a powerful flywheel. The more of our products customers use, the better our data and predictive analytics become. This helps us deliver better results for the students and a more comprehensive view of a student's learning and performance to educators.

This suite of solutions delivers measurable results for students. So students who used Sherpath with our adaptive quizzing products scored higher on our predictive exams more than those who did not. And students who score above our testing benchmarks on our predictive exams passed the licensure exam on their first attempt 98% of the time. Higher pass rates help nursing schools attract more students, more faculty and more funding. Equally important, 3/4 of students who use our simulations improve their clinical judgment skills as they prepare for practice.

Sherpath is our leading courseware solution, which faculty use to deliver content and assignments. And 2 years ago, we launched Sherpath AI. So we will continue to build Sherpath AI to be an agentic learning tool that utilizes user data, context and analytics to anticipate user needs across precise, personalized nursing education use cases.

Let me walk you through a demo. First, let's experience Sherpath AI as an instructor. Now most nursing instructors are practicing nurses who then became teachers. So they understand the nursing, but pedagogy, assessment writing and lesson design are some of the areas where Sherpath AI acts as their teaching partner. Sherpath AI is purpose-built for learning. And the first example of this is that an educator could start with 1 of these 3 prompts based around their typical needs instead of starting a new conversation from scratch.

So let's plan a lesson. Within seconds, we will receive learning objectives, key concepts and activities, all grounded in Elsevier content. But it's domain expertise that calibrates the AI to where these students are in their program. Sherpath AI is built by nurses and educators who know how nursing is taught and tested. Hours of work just became an effortless starting point for lesson planning, just as one example.

Now on the student experience, students' goals are to succeed in their courses, pass the licensure exam and become practice-ready. Licensure preparation is where Elsevier excels. Sherpath AI also has 3 starter prompts tailored to how students actually learn. Students can even filter by the books that they use in class so that the chat's answers stay aligned with their course. Content is where we have a real advantage. Answers draw on our signature textbooks such as Potter's Fundamentals of Nursing, which is the standard for nursing students, not the open Internet, and every answer cites its source on the page.

Now let's break down the concept of how type 1 diabetes affects insulin production. The response is authoritative and accurate because it's sourced from our trusted content. Each citation links to the exact chapter, section, sentence of the e-book where the response was sourced from. And as we scroll down, you'll see that responses are supported by our Osmosis videos as well as relevant images directly from our textbooks, explanation, video and supporting content from one query, all of it is uniquely Elsevier's.

As perhaps one of the best examples of how Sherpath AI is purpose-built for health care education and differentiated, we've designed learning tools directly into the application. To show an example of this, I can request Flashcards to help me study this topic from follow-up questions. I can take the content with me or work through the flashcards directly in the app. And it's the same thing for quizzes. And analytics is where all of this compounds. Ours span Sherpath, Adaptive Quizzing and HESI. So a quiz tells us where a student struggles and what will help. Concept understanding, flash cards, quizzes, study plans and more. The study loop flows without ever leaving the product.

The students will often paste the quiz question in before answering. Sherpath AI recognizes the format and doesn't hand over the answer. The student works through what each option represents first and only then can they reveal the answers the evidence most supports. That's responsible AI built into our product behavior. Now so far, this demo has been showing what we call Study Mode. Tutor Mode moves from ask and answer to a socratic lesson. It gives the student something to think through, then turns it back with a question, all grounded in Elsevier clinical content. Again, concept, video, flashcards, practice and guided reasoning, all in one session.

We are now integrating our AI chat and capabilities across all of our learning systems, including Sherpath, personalized learning plans and adaptive quizzing for a more connected and consistent experience and to increase value for our customers. Each integration reaches a student at the moment they struggle. That signal improves the next recommendation, which drives more use. It's a flywheel. The increasing value comes from the combination of educational content with proprietary usage data sets and sophisticated predictive analytics and a suite of products that cover all the critical use cases across the learning journey, leveraging evolving technologies.

Our road map is about further value enhancement, context-aware AI that gets to know the learner combined with our leading trusted content. So Sherpath AI is a core component of our nursing education offering, and we have seen strong adoption. In 2026, we've already hit 100,000 students, 20% of all nursing students in the United States. As customers see strong value, they buy more nursing solutions. Institutions are choosing Elsevier's products to support all student learning needs across the program. So we are seeing a steady increase in sales.

Here is an example. A leading U.S. university used a limited set of our products course by course with no Elsevier contract, and they spent roughly $200,000 with us in 2025. Now after seeing strong value, they moved to buying the full suite of our modular products on a 3-year contract, which is expected to bring in around $1.7 million in revenue next year. This growth does not come from price increases. It comes from expanded relationships as customers see greater value and better outcomes from using a combination of Elsevier products.

In our top tier of expanded relationships, we see spend per student more than double. I hope that gives you an overview of the nursing education suite of products. The other half of revenues here comes from solutions for nursing practitioners. I'm not going to cover that today, but we have a similar approach there and have recently launched a specialized nursing AI solution that supports nurses in hospitals at the point of care.

And with that, I'll turn it back to Kumsal.

Thank you, Paul. I was watching a Sherpath AI session with a student a few months ago, and at the end, he said, "I love your, Sherpath AI."

Okay. Let me now summarize the key points for decision support. Growth in global research drives increased demand for trusted evidence-based tools. Our strategy is to increase the value of our decision support tools by improving existing use cases, expanding into new use cases and embedding where customers work. And as we add more customer value, there are 4 primary ways we translate that into faster revenue growth over time.

First is adding AI to existing high-value tools such as Scopus AI, usually for a double-digit spend uplift. Second is launching new products such as LeapSpace and driving new business. Third is building interoperable suites such as the nursing education suite that Paul just shared that leads to higher take-up of component tools within the suite. And then lastly, it's integrating into customers' own agentic workflows in a machine-to-machine way, priced on value similar to our risk business.

Our objective for decision support is to continue to improve customer value and to continue to deliver gradual improvements in our long-term revenue growth trajectory with the increasing pace of product innovation. We drive innovation through the combination of our 4 key capabilities: domain expertise, leading content and data, sophisticated linking and analytics as well as advanced modular technology platforms. Now each of these capabilities are critical, but it's the combination that actually creates the customer value.

And now I'm going to hand it over to Jill, who will bring these capabilities to life.

Thanks, Kumsal. All right, the fun part. Okay. Good afternoon. Thanks, everyone, for joining us. My name is Jill Luber, I'm the Chief Technology Officer at Elsevier. I joined Elsevier almost 5 years ago, and that is after spending 20 years in technology at our Risk division. I worked in our data engineering and entity resolution space where I helped create identity linking technology, which allows us to accurately link billions of data sources. Now I work closely in partnership with our Chief Product Officer, Jamie Buckley, who after a decade with Legal, also took an STM responsibility over a year ago.

So I'm going to start with the role technology plays at RELX. You saw this slide in a May risk seminar presented by Vijay Raghavan. I want to start here to reiterate that technology is a key enabler of innovation at STM. At RELX, we have more than 12,000 technologists worldwide. RELX spends $2 billion on technology annually, and we're able to leverage these resources, capabilities and infrastructure at STM. We help build new products as well as data and technology platforms that fuel all data and analytics solutions.

We ensure these platforms are reliable, scalable and secure. We use machine learning and AI capabilities to improve our offerings and enhance customer value at scale. We extract context and insights at scale from scientific content. Billions of records of structured and unstructured content come into our infrastructure. We use AI and other advanced methods to extract entities and relationships. It is not unlike what we do at risk, but it is more complex. Most of our data is unstructured, and we have dozens of entity types.

It's a number that continues to grow. This creates a powerful evidence-based layer on which to apply generative AI. So now I'm going to highlight some of the key capabilities that are driving customer value.

I'm going to start by zooming into our content and data layer for a minute. We have the leading scientific and medical knowledge base, proprietary content of the highest quality as well as decades of aggregating, licensing and curating highly technical databases. We extract billions of data points from that content, citations, authors, research concepts, chemical compounds, substance reactions, tables, images, and all of this creates a rich layer of derived data that fuels our analytics solutions. We then tag, link and validate that data using expert-informed taxonomies, ontologies, so our entities are normalized, disambiguated and put into context. And that's what matters in science and health.

We then build semantic relationships across disciplines to connect concepts, evidence and knowledge. That connected context feeds our multi-model architecture. Specialized agents handle discovery, synthesis, analysis and decision support. It's orchestrated across workflows with human validation, transparency and feedback built in. We've built foundations that assume constant change in enabling technologies in models, in retrieval and in agents as well as change in the customer expectations. So we keep improving without starting again each time as the technology moves. The impact for customers is better outcomes, outputs that are more accurate, relevant and verified.

Let me drill down just a bit deeper on how this works and why it's important in a generative AI world. Generative AI is probabilistic. Our customers need answers which are deterministic. They operate in regulated and critical areas, scientific research, drug discovery, patient care. A single compound may have multiple names. Almost identical names can be completely unrelated. A single data point has different relevance depending on the context. It could be toxic in one context and not in another.

AI agents can answer science questions confidently, but a scientist can only act on an answer they can check. Being probably right and different every time you run a query is just not good enough. The answers need to be accurate, verifiable and have clear provenance. We add curation, standardization and traceability that makes the answer checkable. So the slide shows how we do that. Let's take a chemistry or biomedical question. First, we resolve. We align identities, synonyms, registry numbers and nomenclature. So the system knows when 3 compounds are actually the same name.

Then we retrieve the relevant chemistry, pharmacokinetics, safety and regulatory evidence from our specialized databases, Reaxys, PharmaPendium and Embase. So the data always comes into context. Then we harmonize. We normalize units, parameters, ontologies, protocols and study context. Finally, we deliver structured analysis-ready content and data. We deliver evidence, not answers. We don't decide whether a compound is safe. We make the evidence easier to inspect and leave the judgment to the scientist.

The customer value here is consistency, transparency and traceability. One scientific task delivered the same way every time. And we deliver this value in workspaces such as LeapSpace as well as point solutions. The same approach is used when we embed into customer environments. As an example, we support our sophisticated corporate customers who use language machine learning, AI to build high-value applications, knowledge graphs, synthesis planning, toxicity predictions, integrated into their systems similar to risk. This allows customers to combine their proprietary data with ours, unlocking new customer value.

As corporate customers adopt generative AI and build agentic platforms, this becomes even more important. A recent webinar, a leading pharma customer described how using our solution allowed them to achieve 25% to 30% better quality and precision. Higher precision also means lower costs. So they were able to reduce their context token cost by 30%. I spoke about the importance of our content and data, and now let me bring in domain expertise and our AI approach to complete the picture.

For the quality science demands, a simple retrieve and summarize approach is just not enough. We need to work in 3 stages. First, we understand the question, its intent, concepts and constraints. We retrieve using lexical, semantic and structured search. This includes citation graph and ontologies. We search across text, abstracts, tables, figures and equations. We can then answer with claim to source citations that explain why the evidence was chosen. That explainability and repeatability come from our taxonomies and databases, not from the model. The same approach is applied to both insights and tasks that need an evidence-based approach.

The next step is agentic. Rather than one pass, the system plans and splits the question into sub-questions. Retrieves across the data sources scores its own confidence, names the gaps and can requery up to 3 rounds. It will not respond until confidence thresholds are met. If the evidence is inconclusive, that will be transparent to the user. In scientific and health use cases, an honest but inconclusive answer is better than a confident, but oversimplified one. This is a deliberate design choice.

We build on frontier models from several suppliers, including Anthropic, Google and OpenAI as well as other open source and open weight models like Mistral. Models move fast. Providers are leapfrogging each other and prices shift. So we've architected for change. This week, we have more than 120 approved models across the major providers and choose the right one for each task on quality, latency and cost. This slide shows the loop. We benchmark measuring quality, latency and cost across suppliers, model families and hosting environments.

We optimize. Prompts and configurations are generated for each model family, and they are versioned, traceable and repeatable. The prompt layer is where our domain knowledge is encoded, and we monitor in production with observability, traffic routing by customer, product. And if there's an outage, we fall over to a backup model. Drift and regression feed the next benchmark and the loop reruns automatically when a new model is released. The business point is that the model choice becomes a commercial decision rather than a technical dependency. And switching suppliers is a configuration change, not a rebuild.

On Sherpath AI, model and prompt optimization decreased token costs by nearly 50% this year, while usage and users had a strong increase. We also precompute based on our knowledge graphs and already extracted data points. For example, LeapSpace's Claim Radar is a pre-compute feature. That enables us to manage cost of tokens and have predictable costs for our solutions. We pair this with expert human evaluation. This year, we have already run more than 20 evaluation cycles with real domain experts checking outputs. We have a hybrid approach of automated benchmarks and human in the loop. And that's why we talk about precision AI, not just AI.

Technology also helps us operate more efficiently, which is how we manage cost growth below revenue growth. We've done this through continuous process innovation and increasing the use of AI is speeding this up. In technology, AI-powered spec-driven development is what improves our code quality and shortens our development cycles. Our reusable modular platform is AI-enabled, and we launch products faster while improving our security posture. Over the last 5 years, I've seen a significant improvement in our agility, and we will continue to accelerate going forward with new processes, methods and technologies evolve.

In operations, automated reusable content pipelines let us process growing volumes at lower costs. AI-assisted evaluations of outputs keep accuracy and quality high as we scale. In customer support, AI agents improve response speed and consistency. AI-driven case routing and prioritization make the team more efficient.

In sales and marketing, AI analyzes customer conversations to surface insights and improve conversion and also supports the right channel mix. In internal use, we take the same multi-model approach that we developed for our customer solutions, giving us the ability to optimize quality, speed and cost.

To summarize, there are 4 things I hope you walk away with. First, we are innovating and developing products faster. We build once and we reuse everywhere. Second, we provide evidence-based answers. Precision AI, Trust Cards and expert evaluation keep outputs more accurate, traceable and consistent. Third, our technology flexibility. Our agnostic and multi-model approach means no lock-in, and we move to the most appropriate models as the market evolves. Fourth is cost efficiency. AI-assisted development, pre-computed content and prompt optimization reduce the cost to build, run and support our products.

I'm going to hand back to Kumsal.

Thank you, Jill. That was the fun part. Okay. I'd like to leave you with a few key takeaways as we wrap up. We operate in large global segments with attractive structural long-term growth drivers. We have segment-leading position serving our customers for decades and in some cases, for centuries with critical and complex use cases. We deliver increasing customer value through the unique combination of domain expertise, leading scientific and medical content, sophisticated linking and analytics as well as advanced technology platform.

Our objectives continue to be to deliver higher customer value, improving long-term revenue growth trajectory and manage profit growth ahead of revenue growth. Thank you very much for your time today. I hope that my team and I have highlighted the exciting progress of the business since we last did a deep dive 4 years ago as well as the significant opportunity we see ahead of us for further progress.

We now look forward to your questions, and Paul has already joined me to help with the Q&A.

Great. Thank you, Kumsal. We'll now go to questions. I think for those online, if you can please write your question in the box provided. But I think we'll start in the room. But if you'd like to ask a question, you need to pick up the microphone, which is buried in the seats beside you or in front of you. You must press the button otherwise you won't be heard. And if you could please give your name and organization. And then if you can get all that, who would like to go first?

Right. We'll go with -- we'll start on this side, Joe.

Joe from UBS. Hopefully, I can show people how to use the microphone. So firstly, just kicking off on LeapSpace. Obviously, you're seeing very rapid usage growth, but its revenue contribution is obviously not yet material. So what needs to happen for the current product adoption to flow through as a material revenue driver? Is it just time and the sales cycles as you alluded to? And when can we expect that to be more material as a factor?

And then secondly, when you negotiate prices with your customers around analytics and decision tools, I assume the ROI improvements are a major driver of that. But can you help us understand how those conversations go and where the funding comes from? I guess my point here is this isn't like law where the marginal hour of a lawyer has a value. So how does that change that dynamic?

Great. Thank you very much. You have 2 questions. The first one is LeapSpace user and revenue trajectory. And the second one is how we actually show ROI on our analytics products. I'll take the second one, and I'll ask the first -- I'll ask Max actually to take the first one.

In terms of the ROI, I think we have highlighted to you today with the LeapSpace illustration as well as our nursing suite illustrations, how the ROI differs by customer segment. ROI also differs on whether it's an academic and government customer or a corporate customer. Some of the ROI we deliver is measurable. And Jill, for example, alluded to an example with a corporate R&D customer who -- where we actually integrate our capabilities to their agentic environment, and they could see 30% improvement in precision as well as 30% decrease in token costs. So some of it is really measurable. Others are not measurable. It's time saved as well as quality of outputs improving.

So in each case, I think what we do is continue to show the customer value through better answers, better outputs, faster process and lower cost to serve. And the ROI is sometimes measurable, sometimes it's not measurable, but actually obvious. And based on that, we price the product.

Max, do you want to answer the question on LeapSpace? Thank you.

So Joe, on your question about LeapSpace and adoption and engagement, there's a couple of ways that we think about it. So firstly, we have 1,000 -- almost 1,000 customers across corporate and academic and government using LeapSpace, and we're driving rapid adoption in those customers, good engagement in those customers. Then the second thing is the awareness point. So we're using our touch points in our editorial systems and ScienceDirect. There's millions of researchers interacting with our surface. And we're using those touch points to surface the value of LeapSpace to researchers to let them taste the value, so they're aware of it. And we've got more than 1 million researchers that are now aware of LeapSpace, and we'll keep driving that activation. That's our objective.

The other thing to say is in -- I think it was June, July, Eric talked about the monthly active users doubling in the 90-day period between March and June. And we've seen the kind of new monthly active user addition every month continue at a similar rate, if not a bit quicker. So again, the feedback from customers is extremely positive. The engagement is robust and rapid, both across institutions and awareness. The thing I did say is that an academic cycle can take 12 months, and you need 1 or 2 of those cycles to really have any confidence in the patterns and the trends. So that's kind of where we are at the moment. Very excited about the potential, but still very early days.

Thank you, Max. So to summarize, rapid awareness building and then that actually means more customers, paid customers, and I think it will be gradual, which is usually how our segments behave.

A couple of questions from me, Ciaran Donnelly from Citi. One, just on the decision support tools market size, you talked about corporates and academics. If you could just give us a sense of maybe I know it's nascent, but maybe market sizing, market growth, just to help us contextualize the opportunity there.

And then separately, again, on decision support tools, you talked about the use of proprietary and licensed data. Could you just give us a broad split across the decision support tools in terms of use of proprietary and licensed data?

Okay. So your first question is decision support market size. And the second one is decision support proprietary versus licensed data mix. I'll take both, but my wonderful team here will jump in if I missed anything.

In terms of decision support market size, I think what makes it a little tricky is we have a very large portfolio of decision support tools. You've seen that. It's about 40% of our revenue. About 50% of that revenue comes from research solutions, about 1/4 comes from nursing solutions. Another 1/4 comes from medical education, patient engagement, physician point of care. So the market sizes really differ based on which segment we're serving.

I think in all of these cases, what we're finding with our decision support tools is when we add customer value, we're able to actually translate that into revenue growth. And we do that in 4 primary ways. One is when we add AI capabilities on our existing high-value decision support tools, we're usually able to get an upsell. And the second one is when we launch a brand-new product like LeapSpace, that actually generates new business for us that turns into revenue over time.

The third one is when we actually bring together a suite of solutions similar to our nursing education business, where the suite is more adoption of the individual component products. And then the last one is integrating our capabilities into corporate agentic environment. So that's the way we think about delivering customer value and turning that into a higher revenue growth trajectory instead of sort of market sizing. So hopefully, that helps a little bit.

In terms of our decision support tools and how much of it is proprietary versus licensed data, it's really a mix based on which segment that we focus on. So if you focus on nursing education suite that we showed to you, that's all proprietary. That's all our content. If you look at LeapSpace, that's a combination of proprietary and licensed content, both through Scopus database with abstracts as well as full text articles that we have from more than 1,500 publishers so far.

I think it's really equally important is not just the proprietary and licensed content combination. It's actually the insights we drive from them by extracting billions of data points and linking them through our knowledge graphs, our taxonomies and our ontologies, which creates an evidence layer. That's also an equally valuable layer. Thank you.

Nick Dempsey from Barclays. So I've got 3 questions, please. So first of all, if your Primary Research is growing mid-single digit, then I'm kind of backing out 7% to 8% or so for the rest of the decision support, the other 40%. Has that stayed the same in recent years? Or is it slowing a bit as it gets bigger?

Second question, I guess with the abstract in Scopus, you had originally permission from all the other publishers to create abstracts of their articles and it's your content. But now that you're getting extra value out of that in something like LeapSpace, is there a chance you have to go back to all the other publishers in order to renegotiate that? Or will they want to do that? Because it seems to me a different layer of value now.

Third question, across the different areas of decision support, are you starting to see new AI native competitors or adjacent players emerging? So are your customers having to split their budgets across a lot of new offerings as well as yours even if those don't replace you?

Great. So 3 questions. Primary Research and decision support growth. Scopus abstract values as we use abstracts across new products. And then the last one was about new entrants and competitors. I'll take the first and the third, and Max, I'll ask you to take the second, if that's okay.

So let's start with our growth profile. When we actually presented 4 years ago, we said that our growth had improved from 2% to 4%. And now our growth has improved from 4% to 6%. That's driven by improvement in both Primary Research as well as our decision support portfolio. If you look at Primary Research, it's really high-quality volume growth that actually delivers growth there. And our objective is to continue to gradually increase the long-term revenue growth trajectory in Primary Research as volumes continue to increase.

On decision support, when we presented to you 4 years ago, we said decision support was growing in -- sorry, going back to Primary Research, when we actually presented 4 years ago, we said Primary Research was growing in low single digits. And today, we said it's actually growing in mid-single digits.

Going to decision support, when we presented to you 4 years ago, we said that growth was in mid- to high single digits. And today, we said decision support portfolio is actually growing strongly. And our objective for decision support portfolio is also to continue to gradually improve the revenue growth trajectory there. So if you look at that improvement from 4% to 6%, it's come actually improvement from both our Primary Research as well our decision support revenue base.

I'll answer your third question, and then I'll give it to Max on Scopus. In terms of new AI competitors, I think there is a space for a lot of different products in the segments that we serve. There have always been lots of different providers, including start-ups, traditional competitors, and we see that today as well. We have not seen that impact our customer spend with us so far.

Maybe I'll answer the licensing question in a concrete example of LeapSpace. So in LeapSpace, we have 2 primary data sets that we're effectively building the value on. One is the full text content data that Elsevier has and also other publishers, but also the Scopus data. And we basically have continued to expand the number of -- the amount of content that we license from other publishers. Where the conversations are is really licensing the full text from other publishers in order to use it in LeapSpace. And since we started, Elsevier's share of content is around 20% and its share of citation, which is a measure of impact in science is around 30%.

In LeapSpace right now, we have content that represents 56% of the impact across science right now. And we've been doing that by licensing content, full text content from publishers. And again, we'll continue to have active discussions with publishers and bringing content into LeapSpace. And our goal and our objective across RELX and also across our solutions is to grow revenue ahead of costs and grow operating margin ahead of revenue. So that's kind of how we think about that.

I think it's important to say that we always want to deliver value to our content partners. So we've been doing that with Scopus for a number of years. And now we're actually delivering more value for their content through LeapSpace, which means more visibility, more usage, more people actually using the content in their scientific research.

Just going to go to a couple of questions online before we carry on in the room.

The first question is -- or the first topic really is about if we can talk more about evolving customer value in Primary Research and the sustainability of growth and how quickly we can improve growth in Primary Research, particularly as the model continues to migrate to open access.

And then another question or questions around our approach to M&A in light of discussions around certain assets potentially becoming available? And would it change the competitive dynamic if consolidation happened elsewhere?

Great. Excellent. So I am going to give the evolving customer value in Primary Research to Gemma, and I'll start with approach to M&A. And I'm sorry, I keep changing glasses because I'm at that age where I can't see far, I can't see close. So I need all kinds of glasses.

So our approach to M&A, as you know, at RELX, our strategy is through primary organic development. That's always our preferred methodology. And our strategy at Elsevier is also through organic development as the primary vehicle. In our Primary Research business, from time to time, we'll do small acquisitions to actually fill a gap in a particular discipline or a particular tier, but our strategy continues to be driven very much by Primary Research.

In terms of whether further consolidation would change anything from an Elsevier perspective, I would say though, because we have really large scale. Scale is important in Primary Research, and we already have very large scale. Primary research, over to you, Gemma.

Thank you. So we think about customer value in primary research, there are a couple of lenses which are mutually reinforcing. So if you think about the researcher and what they care about, the researcher cares about being able to submit to a journal that they can trust, that delivers high quality that provides a home for sound science. We also think about that and the ways in which we're continuing to enhance the author experience, the way we're building out our journal portfolio, the quality differentiation between us and other publishers continuing to deliver value there. But of course, that piece around research integrity and the importance of research integrity.

And we talked about the investments that we're making in screening papers pre-submission all the way through to publication and that importance of both quality and trust to deliver value to researchers. And then in turn, the flip side of that for our customers who purchase content from us, who license content from us that's around being able to feel confident that they're accessing the highest quality research that they can trust and have confidence in going forward.

Thank you very much. And maybe to add on to Gemma, as you know, we've shown you today that R&D spend has consistently grown at 5% for decades. Number of researchers have consistently grown at 5% for decades and researcher productivity continues to increase every time there's actually a significant technology improvement. So all of those drive more research, more findings, more knowledge, also the need to be able to understand that knowledge more quickly and to stay on top of it. So that drives both volume in our published articles numbers in Primary Research as well as demand for decision support tools where you want to be able to apply that knowledge into the different scientific cases you're working on, whether that's in a corporate setting or an academic and government setting.

Will, and then Steve.

It's Will Packer from BNP Paribas. A couple for me, please. Firstly, looking -- the performance of the decision support was very impressive looking backwards. But going forward, are we entering a new era as the AI labs verticalize with products such as Claude for Science, which will compete for spend there? And could that weigh on growth? Or on the other hand, could you help us think through the extent to which your view is that they won't enter the space or that your products will be superior?

And then secondly, like other info services providers, you're talking to content differentiation as an important moat. For STM, there's a particular type of content, right? Does the history of political scrutiny of STM where governments and universities heavily fund both the research and the journals bring a different dynamic. And so stopping the integration of journal content with third parties will be more challenging. Any perspective there would be very helpful.

Yes. So I'll -- I think I'm going to take both of your questions and combine them and the team can actually add anything I miss.

In decision support, the customer value we deliver really is actually about combining 4 key capabilities that we have. The first one is domain expertise, where we have real, real deep ties to scientific and health communities, and we understand their day-to-day work. And that's what enables us to actually deliver use cases that really are applicable to those customer segments. So Trust Cards and Claim Radar are examples of that. That's really important.

The second one is content and the breadth and the depth of our content. And we have, again, decades of having proprietary content as well as licensed and curated content, and we have the largest scientific and medical content base in the world. We're leading in that case. The third, which is equally important, I hope that came across in Jill's presentation, is the analytics we've built on this content base over decades. So that's extracting billions of data records from the scientific and medical content base and really creating an evidence layer.

And then the last is our modular approach technology, which is technology agnostic. So when we focus on use cases, all of these capabilities are critical, but it's the combination of the 4 that enables us to deliver customer value in science and medicine where we can actually give better answers, better transparency, better consistency, better repeatability, better reproducibility as well as a lower cost to serve with predictable subscription pricing because we're able to pre-compute a lot of different use cases and also optimize our prompts based on the domain expertise. So that's really the value that we're delivering. LLMs, I think, and AI has been really great for science, and they're going to continue to actually deliver more science activity. And that's a good thing for us as well for our business, that's a driver.

In terms of content differentiation, we have both proprietary as well as licensed third-party and open access content. So when content is available for different use cases, we're also able to take advantage of that. But it's not just the content aggregation, it's the sophisticated analytics layer that we've built on it that actually has created the linking across all the different -- whether it's chemical reactions, whether it is claims, whether it is assertions, all of that sophisticated linking actually is what drives a major part of the value as well. So I hope that gives you a sense.

Saw a few more questions. So Steve?

Steve Liechti from Deutsche Numis. Just on open access, can you just talk about that in terms of the growth rate of Primary Research in the last 4 years? So how much of that growth is the shift from subscription to open access? How much of that is your ability to leverage article processing charges? And I just wonder if that might flip the other way, and we've seen some pressure from the U.S. commentary in terms of article processing charges there. So anything there would be useful. You may or may not want to answer this next one. But what's your definition of strongly in numbers?

So in terms of definition of strongly in numbers, I will leave that to our wonderful Investor Relations team that can give you the lexicon, I'm sure, over drinks. In terms of the Primary Research growth, I think what I'm going to ask is Gemma to help with that question and tell you how Primary Research growth is actually driven in our business.

So open access makes up just about 1/3 of our articles today, but we are business model agnostic. What's important to the researchers that submit to us is that they can trust in the quality and rigor that we provide and what's important to the customers who purchase from us is that they're buying a corpus of content equally that they can trust in and have quality around. And we work with our customers on a range of models, pay-to-read, pay-to-publish, a combination of both.

To the question on different things that happen in different markets, be they policy or regulatory, we work in 100 or 170 countries. So in any given year, there's inevitably different things that pop up, but we have always been really well placed to navigate those and don't see any issues with navigating those going forward.

Thank you, Gemma.

Henry and then, Josh.

It's Henry Hayden from Rothschild & Co. Redburn. Two questions from me today. The first is on the corporate opportunity. So it seems very attractive given the quantum within R&D expenditure and the growth profile there, but it still represents a relatively small portion of your revenues. So how do you think about the positioning of the portfolio to address that market? Are there any shifts that would be needed to tap into it more aggressively?

And then the second question is on Primary Research, sorry. So you mentioned the objective to grow publish articles ahead of the market, which seems to have been working particularly across the last few years. Is this primarily driven by flight to quality? And if so, how do you think about the durability of those market share gains?

Thank you. I'm going to give the flight to quality question to Gemma in Primary Research, and then I will take the corporate question you had.

Corporate is a very attractive opportunity for us. And I think I mentioned today that within our Research Solutions, about 1/3 of the revenues are coming from corporate market. Our strategy is to continue to pursue that opportunity organically primarily. But if we do find acquisition opportunities that are good add-ons that actually fill our strategy, as always, we would consider those opportunities as well. But that's really primarily an organic strategy for us in corporate.

Gemma, over to you for Primary Research.

So our published article growth is ahead of the industry, but you'll notice also that it's below our submissions growth, and that's really by design because we have that rigorous process in place and are very selective and have things in place to make sure that authors can feel confident and comfortable that when they're publishing with us, there is quality and integrity there. We talked about the fact that with the rise of AI and generative AI bringing many benefits for science, there is also that increase in research integrity issues, AI slop.

There's potentially there the opportunity for all of us to think even more carefully about which journals they choose, and they'll look at markers like our field-weighted citation impact and where we sit. And in terms of the -- thinking about the long term, I think Kumsal talked about the underlying growth drivers of the business, again, in terms of R&D spend, number of researchers, researcher productivity, all of which give us thinking optimistically about the future.

Yes. I think in Primary Research, our strategy is very much around high-quality volume growth at a lower effective price unit. And we have a real long-term view here. Science is a long-term game. So when you work with actually researchers who are serving the scientific community, you need to have a long-term perspective as well.

George?

It's George Webb from Morgan Stanley. I guess two last questions. Firstly, when it comes to embedding AI in your decision support tools, and I appreciate it's decision support, not automated decisions. And you've talked about the different ways you enrich the professional with the different aspects of that to make them or help them make the right decision. But how do you kind of manage the residual liability that AI might represent? And how do you think about who that liability might fall on within that kind of ecosystem of parties? Is it the provider? Is it you? That would be interesting.

And then secondly, just more of a straightforward question. On the Primary Research portfolio, where are you on turnaround times at the moment? And do you have an objective for where you think those can get to over the coming years?

Thank you. I am going to give the turnaround times in Primary Research to Gemma. On decision support, the residual liability is managed in 2 ways. One is actually through our technology approach. So I'm going to ask Jill to talk through how we continue to improve the precision of our answers with the technology approach we have. It's also managed through contractual agreements. So we have been in the business of providing information to researchers and physicians for decades.

And there's always contractual -- in our agreements, there are contractual -- there are I think -- if I can just think of the word, there are contractual clauses that actually make sure that the physician and the researcher is in charge of the decision at the end of the day. So that's sort of a contractual approach. But obviously, we never want to put the physicians or the scientists in that position and to make sure that our precision is incredibly high level, we have a technology approach for that in addition to our content, and then Jill can help you with that, yes.

Sure. So the key to our monitoring precision on our answers is in the harness that we build, coupled with domain expertise and experts who are looking at the results that we provide. So we do -- we couple both automated benchmarks as well as domain-specific and domain expert results. And the way we do it is I talked about this harness where we optimize the prompts because you need to understand the prompt going into a model so that you get the most out of it. And that is optimizing for accuracy is number one, and then latency and cost. And we do that for every model family as well as across any model hosting.

And it's the ability to take the benchmark amongst some subject matter experts and really measure against things like accuracy, completeness, correctness. In our health space, one of the things we measure with actual MDs is would this answer actually do harm to a patient. And that's one of the top -- top benchmarks that we want to make sure we don't drop in our accuracy on would this answer actually do harm. So we're taking that then curated domain-specific benchmark, automating it. So when we make any kind of model change, we're modeling against that benchmark again for completeness, correctness, accuracy, doing no harm.

Thank you, Jill. It's also product design because you've seen the traceability that we provide so that users are always able to check the primary source to make sure that they understand where that information is coming from and be able to calibrate the strength of the evidence and how much they can trust it. So that is it. And then Jill also mentioned that we'd rather give an inconclusive answer versus an oversimplified yes or no answer. So we also address that through our product design as well.

Yes, Gemma.

So speed is important to us because it's important to our researchers. They want their findings to be published and they want that to happen as quickly as possible. But we're really careful to do that while at the same time, upholding those quality and integrity standards that our brands are known for. And some of the things we talked about are the investments that we're making in the peer review experience in the author submission experience, but also some of the tools that we're going to be rolling out for editors, all of which are designed to save time in editorial decisions, so editors can be freed up to actually look at and do their job, and we can automate the things that can be automated. So it's that balance of speed, but also making sure we do that while maintaining the quality and integrity that our customers appreciate.

Thank you, Gemma. And our objective is absolutely to increase the speed of the process, and we have done that historically as well. And we really look at sort of that submission process, then the editorial triage, then the peer review process and then the production. And each of those processes have different time frames, and we try to cut days out of that consistently.

Not because I'd like to have the last word, but we have had a late flourish online. So I will ask you a couple of questions to finish off.

Okay.

One from some friends in Canada. Will adoption and usage of LeapSpace be broad-based across the research community? Or will it be concentrated at the high end of research? That's the first question.

A number of people have asked us to expand upon the machine-to-machine sort of risk-like integration that you talked about in decision support. Can you provide some examples? And is that a meaningful driver of growth going forward?

And then finally, could you just talk a little bit more about how AI changes science and how we benefit from this changing dynamic? And does that extend the opportunity beyond research for us?

Okay. So I will ask Max to pick up the adoption of LeapSpace question. I'll take the machine-to-machine opportunity in corporate as well as how AI changes science, which is one of my favorite topics and how -- what that means for Elsevier.

Let me start with the machine-to-machine opportunity. That is a very exciting opportunity. It's early days. And I think you're all familiar with our risk business. So the best way that I can describe it to you is we have a lot of different data sets that we have built over the years. And in risk, we have attributes. And in these data sets, we have what we call connectors. So these connectors make it easier to actually integrate some of the AI precision and the data sets we've built in chemical reactions, in different assertions, et cetera, to an agentic environment and improve the precision as well as the cost to serve for our customers' agentic environments. It is early day. It is an exciting application. It's also early days for our corporate customers where they're building these agentic environments and they're using it in their R&D processes. So more to come on that. But in the early days, similar to LeapSpace, we're very excited about the opportunity.

And Jill, would you like to add some more to that?

Yes, please. And I did spend a lot of time in the risk business. And I just wanted to call out because I did have a conversation with a customer CTO. And I was talking about the machine-to-machine opportunities, and I know that this was a customer in the life sciences space that was getting direct data from us, full data feeds because they were subscribing to a journal. And the CTO actually said, it would be more helpful if you gave me, like you said, this -- when you've already got your knowledge context on top of it and your solution came machine-to-machine with that knowledge, that would really help me out. And because I don't want to have to process the data, I don't want to have to build that knowledge graph on top of what you already know. It's more valuable if you can give that to me directly into our workflows.

Great. Thank you very much. So hopefully, that creates some color.

Now in terms of how AI changes science, let me sort of first talk about the business side of it and then talk about AI changing science. From a business perspective, the 4 key capabilities we have, which is the domain expertise, the leading scientific and medical content, the sophisticated analytics and knowledge graphs that we've built on, on this evidence space that actually improves decisions as well as our modular technology approach, I think, equally valuable in a machine-to-machine environment, and we're showcasing that in corporate already as well as in a machine-to-human environment. So from a business perspective, we're excited about AI and the changes that it's going to bring, and we'll continue to serve the use cases that we are able to serve with our asset base going forward as well.

In terms of how AI is changing science, we're really at the early stages of that. And when I talk to scientists around the globe, there's a lot of different thoughts that are going around. The first one I would say is if you look at historical scientific breakthroughs, a lot of them have come from actually betting against conventional wisdom. This goes back to Galileo, which was published by the Elsevier family when he said that it's actually the earth that moves around the sun, and it was quite controversial at the time.

To recent history when 2 Nobel Prize, finally, Nobel Prize-winning scientists said, ulcers are actually created by bacteria, not by stress, and this is sort of the '80s. So a lot of scientific breakthroughs comes from going against the conventional wisdom and whether AI models, which were probabilistically will be able to actually pull that going against the conventional wisdom and drive the breakthroughs is -- remains to be seen.

There is also -- I'm sure you've seen, for example, a frontier model has actually had a great breakthrough in a math problem in one of the Millennium problems. The issue is we still -- even mathematicians, I certainly don't, but even leading mathematicians don't exactly understand how that problem was solved. That's a problem because when a math problem is solved, it's not the solution that actually drives the scientific process. It's actually understanding the steps and being able to apply that to an engineering problem or a material science problem that actually drives scientific progress.

So there's -- we're still at the early days of AI use. It's clear that it actually helps speed up the process. It's very, very helpful in computational areas with speed and efficiency. But I think we're living through a fascinating time. And if AI can actually accelerate science, that will be an incredible benefit to all of humanity. I mean I'll give you an example, statins, which has saved millions of lives. We don't have heart attacks in 50s and 60s anymore because of statins. Statins was actually first discovered the first step of statin discovery from the first discovery to the drug was about 20 years.

And the very first discovery was in a tiny little Japanese lab with a researcher who was really, really interested in mushrooms. And over time, many researchers built on that, and we had statins 2 decades, which now saves millions of lives. If we can actually squeeze that to be 5 to 10 years, it's really exciting time for both the scientific and the medicine community, but it's very early days for the technology. Sorry, I can talk about this, as you can see for a long time. And my team can talk even more about it.

Do you want to add anything, Max? LeapSpace, yes, come on up. Yes. We had the LeapSpace question.

So on LeapSpace, so we're relentlessly focused on driving customer value. And you saw from the demo just 3 examples of tasks that we're trying to improve for researchers. I mean there are hundreds of tasks that people do in LeapSpace. And for each one, we're trying to drive a better outcome faster and in ways that they can calibrate their levels of confidence and have confidence in whatever output we're giving. So that's the first thing to say.

Secondly, as we've kind of launched it and started to drive adoption, what we've seen, as you know, LeapSpace is resonating with corporate researchers and the value is also resonating with academic researchers. And then when we look at things like discipline mix, we see a broad range of disciplines. So the value is resonating with a broad range of disciplines and also different seniorities of researchers. So we see the market potential as broad in terms of the value that we can deliver, in terms of the tasks that we can focus on the disciplines and the segments of the market that we can serve with the value proposition.

Thank you, Max.

Great. Thank you. I think we'll draw formal proceedings to a close. If you do have any follow-up questions, you know where to reach me, James, don't hesitate to reach out. And for those in London, please join us outside for a quick drink. And thank you very much for your interest, and thanks again for coming. Thank you.

Thank you.

Good morning, and welcome to the RELX trading update. I'll now hand over to CFO, Nick Luff. Please go ahead.

Good morning, everyone. Thanks for joining us today. As you may have seen from our press release this morning, we delivered strong underlying revenue growth of 7% in the first 9 months, and we continue to see positive momentum across the group. Our improving long-term growth trajectory continues to be driven by the ongoing shift in business mix towards higher-growth analytics and decision tools that deliver enhanced value to our customers.

Turning to the performance of each business area. In Risk, which contributes about 40% of the group's profit, underlying revenue growth was 8%, in line with the first half. Growth continues to be driven by our deeply embedded AI-enabled analytics and decision tools with over 90% of revenue derived from machine-to-machine interactions. Business services continue to be driven by Financial Crime Compliance and digital Fraud and Identity solutions and strong new sales. Insurance continues to be driven by the further innovation and adoption of contributed databases and market-specific solutions and strong new sales.

In STM, which contributes about 1/3 of the group's profit, underlying revenue growth was 6%, in line with the first half and a step-up from last year as we continue on the improving long-term growth trajectory. Databases, Tools & Electronic Reference continues to be driven by the rollout and adoption of our AI-enabled solutions built on industry-leading trusted content. LeapSpace, our recently launched research-grade AI workspace is receiving very positive customer feedback and is seeing rapid usage growth. Primary Research continues to be driven by volume growth with article submissions continuing to grow very strongly by over 20% year-to-date. You will hear more about the growth opportunity at STM at our investor seminar on November 3.

In Legal, which contributed a little over 10% of the group's profit, underlying revenue growth was 10%, in line with the first half and a step-up from last year. Law Firms and Corporate Legal with double-digit growth continues to be driven by the adoption of our integrated AI-enabled legal research and analytics platform with its comprehensive verified trusted content and continuously expanding Agentic capabilities. In Exhibitions, which also contributed over 10% of the group's profit, underlying revenue growth was 4%. Strong ongoing growth was moderated by the impact of Middle East disruption of around 3 percentage points. We continue to make good progress on value-enhancing digital initiatives for our customers.

To summarize, we delivered strong underlying revenue growth in the first 9 months. The full year outlook is unchanged. We expect another year of strong underlying growth in revenue and adjusted operating profit as well as strong growth in adjusted earnings per share on a constant currency basis.

With that, we are ready for questions.

[Operator Instructions]

Your first question comes from George Webb with Morgan Stanley.

Just a few from my side. Firstly, just on STM -- just STM with LeapSpace maybe to start with, you kind of highlight the rapid usage growth in today's statement. How should we be thinking about the kind of usage versus monetization at this stage just in terms of how you're going to market? I guess you're trying to put these tools into as many people's hands as you can to a certain extent. So just talking about those 2 things would be great. It's the first question.

Second question was just a confirmation. Can you just confirm that the Law Firms and Corporate Legal segment of Legal continues to run at a double-digit growth rate at the 9-month stage year-over-year? And then the third one, just on Exhibitions. When we think about the fourth quarter and the full year in particular, what are the different factors in there? And maybe just also confirm that the rest of the portfolio is running maybe at the 7% to 8%-ish levels that you've previously discussed?

Yes. I'll take those in order, George. LeapSpace, yes, as you say, I mean, I think it's been doing very well. The product has been very well received. The feedback we're getting from customers has been very good, getting very high satisfaction. The researchers are like the outcomes they're getting. It's given them more -- given the good efficiency, saving time. I think they really value the trust that they can place in what they get from the platform. So that is terrific. I think on usage growth at the half year, we talked about the number of users doubling in the 90-day period from March to June. It's actually -- if anything, we're getting a little bit faster than that since then. So that's why we talk about the rapid usage growth.

And to your question, obviously, at this stage, our focus is primarily on user growth and usage adoption -- usage growth is adoption, et cetera. So as a result of that, the impact on revenue is not material at this stage, but I think we're pretty excited about the longer-term opportunity. And if you -- I'm sure you'll be on at the STM seminar in a couple of weeks' time, and we'll tell you more about LeapSpace at that event.

Your second question was on Law Firms & Corporate Legal. And I did, I think, included in my opening remarks, but just to reiterate it, it is continuing to grow at double digit. It's 70% of the division. It's growing a little bit faster than the division and the division as a whole, of course, is at double digits. So yes, continuing double digit. And Exhibitions, yes, look, Exhibitions has established itself as a high-growth business in the last few years with ongoing revenue growth in the sort of 7% to 8% range. And if you look at the portfolio, excluding the impact of the Middle East, that very much continues to be the case. Obviously, there has been that impact both from events in the region and a little bit of effect on and we talked about this in the first half, a little bit from the travel disruption, including Middle East exhibitors coming to travel, particularly travel sector-related events right through to the fourth quarter.

So I think the impact we expect in the fourth quarter would be similar -- for the full year will be similar to what we've seen for the year-to-date. But underlying it, as you say, I think your question goes to the rest of the portfolio continuing to show good ongoing growth.

Yes. Great. And maybe just one last one just on Exhibitions. Arabian travel market being one of those ones, which I guess is quite important in the mix that maybe weighed down on that Q3 result. It looks like that's set to return next May. I mean on one hand, I'm kind of thinking that the base comp, so to speak, as long as it can run, is potentially soft. But wondering how you're kind of thinking about rebooking rates or anything like that?

I think it's too early to comment on that. And look, the key point is the one you asked about is the strong ongoing growth in the portfolio is what matters and one individual event is -- it tends not to be that material. But the key is the strong ongoing growth of the portfolio as a whole.

Your next question comes from Nick Dempsey with Barclays.

I've got a bit of a dodgy voice. apologies if it's slightly harder to hear me than normal. But just 3 for me. First of all, in terms of the shift of legal customers to Lexis+ with Protege from the Lexis+ platform, can you maybe give us an update on how far through that transition is and whether that continues to run a bit faster than you saw with previous platform upgrades? I guess what I'm getting at is when we kind of run out of that upgrade movement?

Second question, as you look at the STM renewal environment into next year for journal subscriptions, is there anything unusual that you're seeing? Or should we expect a pretty typical year there? And then last question, there's been an allegation of bribery from a Chinese academic against one of your Cell Editors-in-Chief, the Cell Cancer Editor-in-Chief. I don't know if there's anything you can say about the investigation there and what you're doing to kind of look into any corruption allegations.

Yes. Let me take it. So the first one on Lexis+ with Protege and adoption. It's very much continued on the same trend, the same curve that we talked about earlier in the year. And I think we've gone over 50% adoption in the U.S. market and a little bit less than that in the rest of the world with later launches. But that trend is very much continued. And continues to be the case that the vast majority of new sales and the big majority of renewals are onto the Lexis+ with Protege platform. And that, of course, is just the starting point and the people adopting the analytics using them gives us the opportunity to keep adding more value and the functionality and capabilities of the platform keep improving.

I think your second question was on STM. And I think you were specifically asking about the outlook for next year. The trends in science are very strong, and you can see it in the submissions volumes that we're receiving, which obviously is the key growth driver for the primary research side of STM. We are, I think, seeing AI accelerate the pace of scientific discovery. It's accelerating the speed at which people are able to submit. And those strong trends mean that that's coming through in the volumes of submissions we're getting. We, of course, are becoming increasingly selective and the rate at which we're publishing isn't going up as fast as the submissions, and that's a deliberate choice as we maintain and enhance the quality of our journal portfolio. But the underlying drivers, the amount of research going on in the world, the amount of speed of submission, et cetera, are very strong. And I don't see anything that looks like it will change that.

And your last question was about publishing ethics and research integrity. Obviously, that's absolutely crucial to the whole scientific publication process and Elsevier's quality and reputation and the reputation of the journal is absolutely key. We are committed to the highest standards of publishing ethics, and we've got well-established procedures to make sure we maintain the quality and the integrity. And of course, concerns around research integrity is one of the reasons authors can choose established trusted brands that have got the track record, which, of course, are the ones that we have. And of course, no process is perfect. And if there are concerns raised, we will investigate them. I won't comment on any individual case, of course, but we would investigate anything. And I would just reiterate that our retraction rate is incredibly low, a fraction of 1%. So the process is robust, and I think people have confidence in the quality of what we publish.

Your next question comes from Steve Liechti with Deutsche Numis.

I've got a few. Just on academic, just taking a very sort of strategic nonspecific view, do you see any merit in adding to or doing M&A in your primary research side because your M&A historically has been much more skewed towards kind of technology and tools. Just is there a kind of scale argument? That's the first question.

Second question is, can you just a couple of numbers on submissions. Are you able to give us an OA submissions number on the 9 months? And also just confirm, I think your articles published 7% at the first half. Is that a similar figure for the 9 months? And then the final one, just a general one actually. Just on virtual AI assistant news and stuff like that, I'm trying to figure out what that means for you. I mean I can see a positive argument in that more AI-led virtual assistant, more activity in digital, more need for verification. But on the other hand, would an AI virtual assistant be verified and therefore, less work for you to do? Just any color there.

Yes. I'll take those in order, Steve. The first one was the M&A in STM as with all of our business, of course, our focus primarily is on organic development. We've seen the volume growth that we're getting. And so there's plenty of drivers of the growth organically. That includes, of course, launching new journals. We launched dozens of new journals every year as we fill in gaps in the portfolio, expand into new disciplines and expanding disciplines, launched new OA journals in particular. And of course, we've already got the scale. We're comfortably the #1 scientific publisher in the world, already got the scale, which gives us the resources to put behind that business, invest in the technology, make sure we've got the scale to the tools using AI, of course, to support the whole publication process, and we'll tell you more about that at the STM seminar in a couple of weeks' time.

As elsewhere in the group, we'd always look at acquisitions as a way of enhancing and accelerating what we do, and we do occasionally buy specific journals or small groups of journals that fit into the portfolio, but the primary focus is on the organic development. Your question -- second question is about open access submissions and so on. They are open access volumes on an autopac basis, those are both submissions and publication volumes are higher than the overall average, and that continues to be the case. And there's been no change in that trend. It is -- both are growing, both pay-to-read and pay-to-publish, just pay-to-publish growing faster, which is what's shifting the mix. The overall published volumes, as you mentioned, I think we said 7% growth at the first half. There's been no change in that sort of pace, maybe a point either way, but no significant change in that -- in the pace at which we're publishing.

And your third question was on agents and the like -- continues by using -- use by consumers. I think it's fair to say that we're still at the early stages of our customers, the banks and the retailers and other customers of ours, I think early stages, them interacting with agents that are acting on behalf of consumers. We're already helping them detect agentic activity and helping them adjust their risk controls so we can help our customer identify when an AI agent is involved, how it's behaving, how it relates to the consumer that's behind the agent and incorporate those insights into the risk decisions, effectively whether to trust the transaction or whether to require more authentication. But I think it's fair to say that it's still pretty early days. What I would say is that agents, of course, the use of agents creates more complexity, potentially higher interaction volumes, we'll see. But that overall complexity and higher -- potentially higher volumes does give us the opportunity to find new ways to add value for the customers.

Your next question comes from Ciaran Donnelly with Citi.

Yes. One last from me. Just on Legal and the competitor space, one of your peers has been increasingly talking about developing capabilities in the legal research space. So I think, one, could you just outline whether you've seen any change in competitive dynamics within that market? And two, just remind us about the difficulty of replicability of the Lexis solution in database.

Yes. Well, clearly, as you know, our whole strategy across the group, including Legal, does start with our differentiated comprehensive content is verified, trusted, continuously updated and then how we leverage the domain expertise we have to combine that content with the analytics and the functionality and the tools. The content set in Legal consists of billions of documents. We've shown you some of this in some of our seminars, but billions of documents, which we've collected and curated over decades. It is continuously updated from over 3,000 courts and tens of thousands of other sources. And within the content set, there's something like 170 billion linkages and connections that are obviously critical to be able to utilize it in the whole content set in an effective way. We enrich through editorial metadata enhancements, use of technology, use of the -- we have 5,000 -- over 5,000 leading practitioners and subject matter experts in the law.

And that's what makes the content set so powerful and so important for -- in the legal research space. Of course, it is possible for others to access specific content sets. For example, the U.S. case law that's currently available publicly. Several players have had that and been collecting that for many years. But that is only a small part of what we do. And for high-value legal research where you need to be very confident that you're making the right arguments, the best arguments that everything you're citing is up to date, hasn't been overturned and you're going into the high-value, high stakes going to court tomorrow, you need that vast trusted curated content set that we have.

And just on any change in competitive dynamics in that space specifically?

I think I mean all I can comment on is the -- is how our business is doing. And you have seen the continued growth rate in our Legal business with its overall double-digit growth and customers overwhelmingly continue to choose our integrated AI platform. And you can see that in the new sales, which I mentioned earlier in the renewals. And that's where they're seeing value. So that's what we see in the marketplace.

Your next question comes from Will Packer with BNPP.

So we're seeing a rapidly evolving legal workflow market with Astra for Law from OpenAI, Gemini for Legal from Google, Claude Legal, et cetera, as well as the developments at Harvey and Legora. So we're seeing quite different strategies from the Clios, the Westlaws, LexisNexises in terms of integrating their research content into that orchestration layer. Lexis from the outside appears to be most cautious on integration in those third-party workflows. Could you help us think through the key strategic considerations as you develop your strategy on integrating with those new workflows, which one would imagine are getting more and more important. On first principles, it would feel that maximum integration will be the client's preference, but perhaps that's wrong. Any color there would be very helpful.

Yes. I mean just to reiterate, obviously, our starting point in this is, as I just mentioned, is the -- and where our strategy is based in is that trusted differentiated comprehensive content set that we were just talking about and adding the functionality, the capabilities and the tools, et cetera, around that. We do that on our own platform. We also do that through partnerships and integrations that we have with others. And we've talked about this in the past. We have something like 25 different partnerships and integrations. They include things like Microsoft, of course, where a lot of work in Microsoft. And we've announced new ones in there just in the last quarter. And of course, we utilize, as you know from previous presentations, how we utilize all the different large language models. I have a deliberate technology-agnostic multi-model approach to make sure we're using the right ones. And our objective is to provide the functionality and interoperability that our customers find valuable.

So we always evaluate every new tool that comes out, every new piece of functionality and assess whether the best thing and the best way to utilize that and provide value to customers in the context of what we do for them, whether it's to replicate that functionality, whether it's to build it into our platform or whether to enable that interoperability between products. Our focus tends to be on the substance on exactly that, what value can we add to the customers, taking feedback from the customers, and we'll continue to take that approach and build in the interoperability and the integrations that we think our customers will find most valuable.

So it sounds like your approach is kind of more pragmatic and agile will be a fair characterization.

Yes. You might have seen just this quarter, we announced the setting up of our innovation lab, which is the -- an environment where we're bringing in legal professionals from law firms, from corporate to work directly with LexisNexis product developers, LexisNexis engineers, taking their feedback real time and working on things that they -- problem areas or opportunities for efficiency that they're feeding back to us and looking at how they work in practice. So it's not about press releases and announcements. It's about real practical tools and functionality and that interoperability that the actual end users find valuable.

Your next question comes from Natale Puccia with TD Cowen.

So obviously, Legal has continued to grow well above its historical rate as AI-enabled products have scaled. Maybe could you just help us understand qualitatively how much of the incremental growth today is being driven by AI-enabled products and higher customer value versus the underlying core research business? And perhaps maybe briefly on STM. I don't want to steal any of your thunder from the upcoming seminar, but I know monetization perhaps is a bit of a longer path. But does the value from these AI tools, do you believe they will show up principally through higher ARPU within existing customers or I suppose, the expansion of the addressable market?

Yes. On your first question about the revenue and how much from AI-enabled. I mean, of course, it's very hard to separate it because this is absolutely integrated into the platform, the core flagship platform in Legal, Lexis+ with Protege, of course, is fundamentally built on the Legal research platform and then adding the Agentic capabilities and the functionality on top of that. I think as I mentioned earlier, the vast majority of new sales and comfortable majority of renewals are on to that platform. As we've discussed in the past, there is an uplift in spend when someone moves to that platform compared to not to the older platform. So obviously, that's an important component of the of the overall growth and the acceleration of the growth we've seen in recent years. But I can't separate it out in the sense of perhaps that you're looking for. It's just an overall value and overall revenue growth being driven by the provision of that value.

Your other question about STM, as you say, I won't steal in terms of thunder for 2 weeks, less than 2 weeks away until the STM seminar. I should talk a bit more about this there. I think it's true that with STM having a more diverse product suite with a broader range of use cases, a broader range of customers. Some of those customers do have longer decision-making cycles. Obviously, it is a large subscription-based business, too. So changes and the benefits of adoption of new tools comes through gradually into growth rates over time, and that's fine. We do see significant opportunity of the application of generative AI and other new technologies in the scientific and health care space and specifically in relation to what we do. We think that creates more value, significantly more value for customers who utilize our products as we build that functionality and capability in. And overall, over time, that value is reflected in our revenues. And that's what is the key growth driver. And as you say, we'll talk about that more at the seminar.

Your next question comes from Adam Berlin with Goldman Sachs.

Two for me, please. The first question is following up on the discussion around news earlier. You said it was a bit early, but one obvious application would be people using news to kind of search car insurance quotes to see if they can search maybe 20, 30 car insurers rather than just 1 or 2 online. Have you seen any impact there from people using that application yet because that might be something you have seen in a couple of months since news has gone live. That's the first question.

The second question is, when you say that Protege platform has got over 50% penetration, do you mean that the number of law firms is more than 50% or the number of users is more than 50%?

Yes. On your first question about insurance, volumes, I think it's fair to say we -- as of yet, there's no noticeable -- bear in mind, the numbers are vast. And we -- all U.S. consumers, all U.S. car owners and drivers. So the number who might be using an agent to do shopping for them does remain a small fraction of that. So at this stage. But as you rightly say, Adam, clearly, there is potential for higher volumes of activity. We have -- the overall insurance market activity hasn't particularly changed. It remains elevated as people -- as the market, we've seen higher volumes in the last couple of years have continued at higher levels but we'll see where it goes from here. Just to reiterate, of course, the key driver for us is the innovation and not necessarily the volumes themselves will vary over time. But it's actually -- for us, it's about the innovation, new functionality, new products, potentially based around and related to the use of Agentic consumers using agents, but we'll see how that evolves.

And your second question was about Lexis+ with Protege. When we talk about the base and being over half now in the U.S. on the new platform, Lexis+ with Protege, we're just waiting that by revenue, the revenue mix of the overall business.

Your next question comes from Henry Hayden with Rothschild & Co.

There's some new language in the release around LeapSpace, which seems from our end, a very positive development. We're wondering if you could share some thoughts around the risk of that solution cannibalizing Scopus AI and ScienceDirect AI given the overlaps in functionality. Secondly, strong new sales in Fraud and ID again. We were hoping you could give us an update on the state of underpenetration in that market, given that this has been a recurring piece of language, particularly against the accelerating backdrop of agentic fraud. So in other words, is the market growing faster than it can be sold into?

And then finally, we're hoping for an update on the difficult comps in auto transactional volumes. Are you seeing those already come through against that 8% growth figure? And how should we think about that trajectory into next year?

Sorry, Henry, I didn't quite capture your second question. What was the topic?

On the state of underlying innovation in Fraud & ID, just is the market growing faster than you can see on the new solutions.

Okay. Your first question about LeapSpace and Scopus and ScienceDirect AI. I will slightly defer to the STM seminar in a couple of weeks' time. The Scopus does a different use case and does slightly different things, and we can talk about a bit more about that at the seminar. And that continues to go well as a product. It's on similar trends to what we saw earlier in the year. On ScienceDirect AI, that product has effectively been wrapped up into LeapSpace. LeapSpace is a successor product to ScienceDirect AI. So users and customers who bought ScienceDirect AI have been migrated to LeapSpace. And it's -- that's the next generation with all the additional functionality and things that come with that. And again, we'll talk about that at the seminar.

On Fraud and Identity, obviously, that's the business services, of which the Fraud and Identity is just over 1/3 of that in terms of that segment within risk alongside Financial Crime Compliance being the other big part of it of that particular segment. It's been an important growth area for a number of years, continues to be. It is where we are seeing lots of opportunity to continue to innovate, continue to bring new features, new functionality. It's where we've moved from identity being about the person and the individual to still being about that, of course, but also being about the device that they're using, also about how they use that device, so-called behavioral biometrics, document authentication, which relates to that, and we talked in the seminar on business services back in May about IDVerse, which is a functionality that we got through an acquisition that we've integrated in the platform.

So continuing to see lots of opportunity. And of course, with Gen AI being used by the bad guys as it were, that does create new challenges for our customers. And when you get synthetic IDs and deep fakes and so on. And therefore, the importance of what we do becomes all the greater, and we continue to see strong demand for the F&I products, Fraud & Identity products with expanding use cases and us adding new functionality as we help customers combat those emerging threats.

And your last question was on auto insurance volumes. I think, again, I'd reiterate that the key growth driver in risk as a whole and insurance is the innovation. It is coming from the introduction and rollout of new products and new features. That continues at pace, and that's what's really driving that consistent high single-digit growth that you're seeing in risk. Of course, we do have some volumetric pricing, particularly in insurance. So the volumes that go to the platform are relevant to us. But they tend to even out across the different segments. Insurance volumes do remain high, shopping and switching activity. They grew quite strongly over the last couple of years. So we're not seeing quite the same growth that we've seen, but they do remain high. But if you combine that with what we're seeing in the business services and look at risk as a whole, the different range of factors that influence the transaction volumes, they do typically bounce out over time, and that's why you've seen the consistency in the overall growth rate driven primarily by that innovation.

This concludes our question-and-answer session. I would now like to turn the conference back over to Nick Luff for any closing remarks.

Okay. Thank you, everyone, for joining us today, and thanks for your questions. And just if I haven't mentioned enough times already, we have a seminar on STM in a couple of weeks' time, which we'll see you at that. Thank you.

Good morning, everybody. Thank you for taking the time to join us today. As you may have seen from our press release this morning, we delivered strong financial results in the first half, we made further operational and strategic progress, and we continue to see positive momentum across the group. Underlying revenue growth was 7%. Underlying adjusted operating profit growth was 9%, and adjusted earnings per share growth was 11% at constant currency. All 4 business areas continue to perform well.

On this chart, you can see the relative sizes of the business areas and their growth rates. Risk with continued strong growth, STM with a step-up to strong growth, Legal with a further step-up in growth, and Exhibitions with continued strong ongoing growth. In Risk, underlying revenue growth was 8% and underlying adjusted operating profit growth was 10%. Strong growth continues to be driven across segments by our deeply embedded AI-enabled analytics and decision tools, leveraging our unique contributory and proprietary data sets with over 90% of revenue coming from machine-to-machine interactions.

In Business Services, which represents over 40% of divisional revenue. Strong growth continues to be driven by Financial Crime Compliance and digital Fraud & Identity solutions and strong new sales. We continue to expand our extensive differentiated data assets with integrated advanced authentication and behavioral intelligence to address the increasing complexity of risk decisioning for our customers.

In Insurance, which represents around 40% of divisional revenue, strong growth continues to be driven by further innovation and adoption of contributory databases and market-specific solutions and strong new sales. We continue to expand our products adding data sources and analytics to enhance value for our customers. For the full year, we expect continued strong underlying revenue growth with underlying adjusted operating profit growth exceeding underlying revenue growth.

In STM, underlying revenue growth was 6%, a step-up from full year 2025, driven by the evolution of the business mix towards higher growth, higher value analytics and decision tools supported by the increasing pace of new product introductions and strong new sales. Underlying adjusted operating profit growth was 8%.

Databases, Tools & Electronic Reference, which represents around 40% of divisional revenue, delivered strong growth, driven by higher value-add solutions with continued rollout, adoption and usage growth of our AI-enabled tools. We continue to expand our solution sets with new releases built on industry-leading trusted content, including our research-grade AI workspace, LeapSpace, which has been positively received by customers.

In Primary Research, which represents a little over half of divisional revenue, good growth continues to be driven by volume growth. Article submissions continue to grow very strongly across the portfolio by over 20% in the first half, with the number of articles published, growing 7%, in line with our long-term average. For the full year, we expect continued strong underlying revenue growth, with underlying adjusted operating profit growth exceeding underlying revenue growth.

In Legal, underlying revenue growth improved further to 10%, driven by the continued shift in business mix towards higher growth, higher value legal analytics and tools supported by strong renewals and strong new sales. Underlying adjusted operating profit growth was ahead of underlying revenue growth at 13%.

In Law Firms & Corporate Legal, which represents around 70% of divisional revenue, double-digit growth is being driven by the continued adoption of Lexis+ with Protege. Our core AI-enabled legal platform with its integrated agentic assistant. Ongoing releases of new functionality and tools integrating additional skills and capabilities into our core platform with its comprehensive verified legal content is increasing the value add of our trusted legal AI. For the full year, we expect continued strong underlying revenue growth, with underlying adjusted operating profit growth exceeding underlying revenue growth.

Exhibitions delivered strong underlying revenue growth of 6%, reflecting the strong ongoing growth profile of our event portfolio, slightly moderated by some travel disruption. First half underlying adjusted operating profit growth of 2% also reflected event cycling, timing and the rescheduling of some events to the second half. We continue to make good progress with our growing range of value-enhancing digital tools. For the full year, aside from uncertainty around remaining events in the Middle East, we continue to expect strong underlying revenue growth with an improvement in adjusted operating margin over the prior full year.

Our strategic direction is unchanged. Our improving long-term growth trajectory continues to be driven by the ongoing shift in business mix for its higher growth analytics and decision tools. This is being supported by the continued evolution of artificial intelligence, which is enabling us to add more value to our customers and to develop and launch products at a faster pace.

Our growth objectives remain for Risk, to sustain strong long-term growth; for STM and Legal, to continue on their improving growth trajectories; and for Exhibitions, to sustain strong long-term growth. When combined with continuous process innovation to manage cost growth below revenue growth, the result is a higher growth profile with strong earnings growth and improving returns.

I will now hand over to Nick Luff, our CFO, who will talk you through our results in more detail. I'll be back afterwards for a quick wrap-up and Q&A.

Thank you, Erik. Good morning, everyone. Let me start by providing more detail on the group's financials. As Erik said, underlying revenue growth was 7% with underlying adjusted operating profit growth ahead of that at 9%. As a result, the adjusted operating margin grew by 70 basis points to 35.5%. Strong operating results flowed through to adjusted earnings per share, which at constant currency increased by 11%. Cash conversion was strong at 98%, and leverage was 2.3x and up slightly from the year-end, reflecting the first half bias of dividend payments and the buyback.

Given the strong financial performance, we are increasing the interim dividend by 7% to 20.9p per share. We spent GBP 103 million of the 2 acquisitions in the first half, and we deployed GBP 1.75 billion out of the plan, GBP 2.25 billion, share buybacks this year. Looking at revenue, you can see here how all 4 business areas contributed [ to the overall ] 7% underlying growth. The group as a whole, total revenue growth at constant currency was 5% after minor portfolio effects in Risk and Legal and after Exhibitions cycling timing and rescheduling effects as well as a further step down in print activities.

In sterling, total revenue growth was 3%, impacted by the relative strength of the pound against the dollar compared to the prior year. Here, you can see the 9% underlying growth in the group adjusted operating profit. We continue to target cost growth to be below revenue growth in each business area, as a result, Risk, STM, and Legal each delivered underlying profit growth 2 or 3 percentage points ahead of underlying revenue growth. Exhibitions growth reflects the event scheduling referred to earlier. After portfolio effects and the decline in profits from print, total adjusted operating profit growth in constant currency was 7%.

There's a similar currency effect on profit as there was a revenue giving adjusted operating profit growth in sterling of 5% with profit growth ahead of revenue growth in Risk, STM, and Legal, margins improved in those business areas, driving over -- improvement of 70 basis points to 35.5%. Exhibitions margins reflect the event rescheduling, but they remain well above historical levels.

Turning to the group adjusted income statement. You can see here the underlying growth of 7% in revenue and 9% in operating profit. The interest expense was almost unchanged with the effect of higher average debt levels, offset by lower average interest rates. The effective tax rate was 22.8%, up 30 basis points on the prior full year and first half. Net profit was up 7% at constant currency and up 5% in sterling to over GBP 1.2 billion. With the lower share count as a result of the share buyback program, adjusted earnings per share were up 11% at constant currency and up 8% in sterling, 68.6p.

Turning to cash flow. Cash conversion was strong at 98%. EBITDA was almost GBP 2 billion and CapEx was GBP 292 million, equating to 6% of revenue. After interest and tax, total free cash flow was over GBP 1.1 billion. And here's how we deployed that free cash flow. We completed 2 small acquisitions in total consideration of GBP 103 million and then 1 small disposal with proceeds of GBP 62 million.

Dividend payments were GBP 851 million. And as I mentioned earlier, we've completed GBP 1.75 billion of share buybacks. Overall, net debt was GBP 8.7 billion at the end of June. Ratio of net debt to EBITDA calculated in U.S. dollars, was 2.3x in the middle of our typical range of 2 to 2.5x.

With that, I will hand you back to Erik.

Thank you, Nick. Just to summarize what we have covered this morning. In the first half, we delivered strong financial results including a step up in growth in both STM and Legal, as we made further operational and strategic progress. We continue to see positive momentum across the group and we expect another year of strong underlying growth in revenue and adjusted operating profit as well as strong growth in adjusted earnings per share on a constant currency basis.

And with that, I think we're ready to go to questions.

[Operator Instructions] Our first question comes from George Webb of Morgan Stanley.

Well done on the first half numbers. I got a few questions, please. Firstly on STM I think you called out that you're still seeing over 20% article submissions growth in the first half. And I think you said publications at 7%. For last year of FY '25, I think it was north of 20% as well and publications of 10%. So just curious what's driven that publication growth moderation and whilst the submissions growth has stayed pretty -- well, still stayed above 20%.

Second question and a bit tied to that, are there any numbers we can think about at this early stage with regards to adoption of LeapSpace into your customer base? And then lastly, for you, Erik on the Legal business, it's nice to see the tick-up in growth there now at 10% underlying. We talked a bit last year about how you see the different dynamics between the legal reference content market and the legal workflow market, just given how quickly the whole space is evolving, both in terms of where certain labs are talking to playing and where to AI natives are growing.

I'd be curious for your latest thoughts on the overall growth opportunity you see there for RELX in the competitive landscape.

Okay. So on STM, yes, we continue to see that there is significant growth in submissions to the company. I think that's a combination of what's going on in the industry and the global science and research market where you see that the number of researchers continue to grow, global research spend continues to grow and productivity tools make researchers more productive. That's why we continue to see that increase. I think there might also be an element here of some of the concerns about research integrity could lead to slightly higher volume for us with a longer history, bigger brand, et cetera, and many other well-known brands in our portfolio that might be a component.

When it comes to the long-term trend, we continue to expect that we will see strong volume growth for many years to come. I don't expect it to be remaining over 20%. I think, historically, we used to say that submissions would grow in the high single digits. At this point in time, we probably would see it moderating down probably to low double digits on average over time, higher than before based on these drivers, but probably not at this level.

When it comes to the amount of publications, the actual articles we publish, as you can see, we have basically all the time on any time period published a fewer and fewer of the articles that we received, meaning we're becoming more and more selective, more and more targeted. And I think that's important in our quality positioning in the industry. So that the number of articles published should grow below the number of articles submission-ing growth over any longer time period. In any 1 short time, it can, of course, be different. But from our perspective, I think this is -- that we are seeing a very rigorous quality-focused selection process in our 3,000 journals. So it's intentional from our side.

The next question you asked was on LeapSpace, right and how we're doing on penetration there. I think it's important to look at LeapSpace as a continued evolution of the ScienceDirect AI that we launched a year earlier but LeapSpace overall is going really well. This has continued to do well if you look at it as an extension of ScienceDirect AI or if you look at it as a stand-alone launch because it's basically so much new functionality is going very, very well.

It's too early to really talk about any penetration curve because of the 2 alternative interpretations you can have as a transition from ScienceDirect AI or a new launch but we can tell you the key things that we focus on is, number one, the customer feedback is very, very positive, very high user satisfaction, very specific comments on how it supports their critical thinking, how it saves them time and in very specific examples that we have studied and asked them.

The second thing is we're seeing a lot of interest from institutions that see that value. And most importantly, we can see that the number of active users has grown significantly since we did the full transition from ScienceDirect AI to LeapSpace. I mean, for example, in the 90-day period from March to June, I think -- so a number of active users are almost double. Now it's very early to look at that and to do any trend line from that because it involves early use, early customers and some trials. But the other very important factor is that the amount of usage during that time grew significantly faster than the number of users, which means that the users see the value and use it more. Those are the early trends that we're seeing, very positive.

Then on Legal, you asked on where we see that going over time. The way we see it right now is that the step-up in growth is driven by the continued development and rollout of higher-value AI-enabled tools that we are putting into our core platform and the things that we put on top. What has happened is that the continued growth, the continued penetration of Scopus, with -- sorry, with Protege. I'm mixing the product here. Lexis+ with Protege. We continue on the same penetration curve, the trend that we've had before in terms of share of revenue. What we can see in terms of number of customers is that it's probably growing slightly faster now in terms of number of institutions. Users is going faster than a number of institutions and usage growth is faster than the number of users. So it's on a very positive trend. And what we can see is that -- our customers really see the value in these tools.

The tools continue to get significantly better on a regular basis. There are upgrades coming probably every couple of months that are significant that our customers see and see more value in. The fact that there are other companies out there that also leverage generative AI tools in order to provide efficiency, processes, workflow tools, and software improvements in the broader legal tech industry, we think is natural and probably good things for the overall understanding of these tools in the industry.

We come from the information-based side, which is the smaller side of that. And we believe that when you put these tools on top of the unique comprehensive broad-based, verified, trusted content sets that we have, the value increase is significant. And we think a large number of our customers are going to continue to use those tools on top of our content sets. That does not mean that, that's at the expense of anybody else that works on the workflow or software technology space, which is significantly larger than the information based tool space that we're in. But we continue to see significant upside to help the legal industry for many years to come and we expect that we will continue to do well and take a share of that upside in terms of how we can add value to our customers.

The next question comes from Nick Dempsey of Barclays.

I've got 3, please. So first of all, you referred to strong new sales in Risk, STM and Legal in the release. As those we know will flow gradually into the revenue growth. Does that give you good confidence on the growth rates for those 3 divisions into next year and even beyond that? Second question, just going back to Legal. Can you talk about how much of your growth is coming from upgrading customers who were not on the Lexis+ AI platform to that platform? And how much from customers who already upgraded to that but are now paying for additional offerings such as linking Protege into internal documents or other workflows on top. Just trying to understand the dynamic there.

And the third question, you've seen good margin improvement in Risk, STM and Legal in the first half more than we've typically seen in the past. Are there any timing effects there? Or can we expect good progress on those margins for the full year?

I'll take the first and I'll ask Nick to cover the third one. Strong new sales, we mentioned that in most of the segments, as you mentioned, and it is a very important factor in the long run in terms of being an indication of the momentum in the business and our customers see our new tools and the value they deliver. It is not the most important factor in driving revenue growth in the current year because new sales is a small part of the overall revenue of the company. And it might not even be a big indicator of where it's going in the next 12 months, but it is an important indicator of the momentum in the business and over time, a driver of the long-term growth trajectory. I think that's the way you need to look at it.

When you talk about Lexis+ with Protege, we are seeing that new sales are increasingly almost exclusively now on Lexis+ with Protege. I think we said last year that the vast majority of new sales value was Lexis+ with Protege and Lexis+ AI and now so far this year, we're running pretty much like 90% of the value of new sales is coming from the AI-enabled platform. And when it comes to renewals, we're -- now up to roughly 3/4 of the renewal value is coming from the Lexis+ with Protege package.

But you asked also what do we see from there. What we then see users that are -- users and customers that are starting to use some of these tools. They brought on them. They want to use more of them. They see more value. And that's the point that I made before about the fact that number of users is growing faster than number of institutions or than value and the usage of different tools is growing faster than users.

So we continue to see an increase in value-add and in users in usage and upside opportunity after they have made the first conversion. We see the first conversion to the core AI-enabled platform as the starting point of the future growth opportunity, not the end point.

And Nick, to your question on margin, as you know, we focus on ensuring that cost growth remains below revenue growth in all of our businesses. That's what's enabling us to drive profit growth to be faster than revenue growth. And as you've seen over the last couple of years, the -- that gap is, in a positive way, has become a little bigger as we've seen revenue acceleration, and we're using GenAI internally in our processes, allowing us more efficiency. So that gap has got a little bit bigger. We absolutely expect that to continue. And of course, all things being equal, that will drive good expansion of margin over time. Exactly where that lands in any 1 period, of course, depends on currency and M&A effects and the like. But all things being equal, we're looking to continue to drive that forward.

The next question comes from Adam Berlin of Goldman Sachs.

I've got 2 questions on the Legal division. The first question, just following up on what you just said, Erik, about Protege. You described it as the upgrade as a starting point in adoption and then you can grow revenues beyond just converting that customer to an AI platform. Can you just give us a bit more detail on kind of what that product look like? What are the additional add-ons customers can buy so that you get more revenue from them after they've actually just moved to the Protege platform? So I just thought that was a one-off benefit. So it sounds like there's more revenue opportunities even after the upgrade. So if you could give more detail on that.

And the second question is, I understand that you provide these Protege add-on with -- on a fixed subscription basis. So you're not passing on the underlying cost of tokens to your clients. Do you need to move to some form of consumption-based pricing for tokens? Is that token usage goes up a lot? Or is there some reason that -- that may not happen. Do you have better economics in terms of token usage or better technology. That means less token usage than some of your competitors. Can you just describe how that's operating in practice?

Sure. I'm going to ask Nick to cover the second in detail. But I think on the first one, I think it's very important that you look at this over a period of time. We keep upgrading and adding tools and functionalities to Lexis+ with Protege on an ongoing basis. It's a significantly more valuable platform and significantly broader and more valuable range of tools today than it was even 2 months ago, and 2 months ago was significantly better than 4 months ago. And I think you have to look at the transition to Lexis+ with Protege, basically the core integrated AI-enabled platform as a necessary starting point for the future growth opportunity for the next decade or so.

I think you have to think of it all more similar to the print to electronic transition. You could not -- we could not start selling analytics and decision tools to somebody who had print books or print journals. You had to transition them to the electronic platform, the basic core electronic information retrieval platform, the reference platform in order to then be able to offer higher value-add analytics and you have to have basic analytics enabling platform before you can start to sell fully integrated sophisticated decision tools.

So we've been on that journey for many years before, moving people on to core AI-enabled integrated platform is the next foundational piece that will then drive revenue for many years to come.

So can you just give us like an example of what you're talking about? Like what's something you've launched in the last couple of months that you're charging for beyond just the upgrade to the platform, just to give us a sense of what kinds of things are in there.

So I mean a single example might be Lex Machina, which is the legal analytics, which brings strategic insights in from historic cases and helps you judge whether you should take a case, how long it might take. We've integrated that into Lexis+ and Protege. We announced that a month or 2 ago. So it's integrating that as just one example of many, many examples.

And Adam, your question about token cost, I think we see managing token costs effectively managing on behalf of customers is a significant competitive advantage because the way we have our technology configured, we can maximize customer value, dynamically managing the use of which model, which agent are you using in which particular use case, which particular step of the process. And because we, of course, have preprocessed all the underlying content in a way that allows us to operate very efficiently. It tends to be less token intensive in its use.

And of course, our customers are wanting us to help them manage their overall token consumption. So I think we're seeing that as a significant competitive advantage. And of course, the cost of tokens is going up, but it's still less than 1% of our overall cost base. And in the context that I was talking about earlier about keeping cost growth below revenue growth, given all the efficiencies that are coming as well, I think that's very manageable within our overall approach of keeping cost growth below revenue growth.

The next question comes from Ciaran Donnelly of Citi.

Just a few questions left for me. Firstly, just in terms of the ability to leverage AI internally, has that reduced the need for M&A given the ability to develop products organically has probably become easier? Two, just from here on STM in terms of the goal to accelerate growth further. Could you just outline what are the key dynamics you need to see or the key initiatives in terms of driving that growth further from here? And then just thirdly, just on the visibility of the scheduled events in the Middle East in the Exhibitions division. Can you just give us some thoughts on how you see those trading?

So maybe I'll ask Nick to cover the first and the third, and I'll cover the second.

Yes, Ciaran, absolutely, the use of GenAI internally is certainly helping us do things faster. It's certainly part of what's enabling us to come faster to market with new innovation, the pace of new product introduction. That's absolutely doing that. And our focus, as always, is on organic development primarily. I mean that's what's the key driver for the growth across the group and the key value creation for shareholders is coming from the organic development. We will continue to look at M&A where we see things that can enhance and accelerate the organic development. And where we're the natural owner of something, we will continue to do that. And what exactly we spend in any period will depend on what happens to come up and what's available. So it will vary from period to period, but -- and we'll still look at that. But absolutely, the key focus is on the organic.

Yes. On STM, the main driver of the growth pickup so far has been the continued development and the rollout -- further rollout of the higher value-add AI-enabled tools. And we think that's exactly what will continue to be the main driver for many years to come. We're really very, very early in the rollout stage of these tools in the science and research industry. I mean, as an example, the big tools that we've had, if you look at Scopus AI, we've talked about for a while, that has continued to do well. It's on the same path as before. It continues to grow. Customers see similar value uplift to what we see in the legal space and the adoption curve has similar shape to legal, but it's a bit slower.

The customer base is more fragmented. The decision cycles take longer. So we've said before it takes sort of 50% longer almost. But it's on that same path, and we continue to see users growing quickly, usage growing more than users, et cetera, exactly the same patterns that we talked about before. We're in the early stages of this. We expect that to continue for several more years. And then we can see that the newer tools that we are developing like LeapSpace, they are similarly adding significant value. And we're just at the early -- very, very early stages there. And we've done this in several other smaller areas around the company, too. We've had slightly different approaches in other parts of the company inside the sort of medical education, SurePath AI, which we talked about before.

We had a slightly different approach to packaging and pricing. But the way we've done that and combining products with the AI capability has led to double-digit growth in that subsegment, but we're still in the relatively early times there. So we think there are a lot of opportunities across Elsevier to continue to build tools that add more value, and that will, therefore, have increased users, increased usage and therefore, improving growth rate. But we do have to remember in STM that it is a more fragmented product suite. It is a more fragmented customer base, both in types of customers and geographically. And they have slightly longer decision cycles in many of those customers. So the growth is likely to come through gradually, but we think there's potential there for it to continue to increase gradually for very many years to come.

And Ciaran, your question about the Middle East and Exhibitions, I mean, just to scale that, they represent about 4% of the divisional revenue, and those left to run perhaps more like 3% of the divisional revenue, so less than 0.5% of the group's revenue. We have indeed rescheduled, as we mentioned in the release, we rescheduled some of those into the second half, we are still planning to run most of them, but we have to acknowledge there is uncertainty around them and exactly how they perform. But I think I'd focus back on the other 97%, if you like, where we're seeing the strong ongoing growth in the overall portfolio, and that's the key focus going forward.

The next question comes from Will Packer of BNP Paribas.

Two for me, please. Firstly, the Trump administration has made a flurry of announcements on scientific research funding in recent weeks, which are making lots of headlines in the trade press. While the proposals are wide-ranging, the most relevant to Elsevier is the OMB proposal to make journal subscription costs and APCs unallowable expenses on federal research awards. From our conversations with those closer to the budget process, we understand these proposals have a real chance of being enacted for 2027.

With 10% of global research federally funded and perhaps 40% of open access funded, could you talk through how these proposals would impact Elsevier if introduced? What tools could Elsevier use to mitigate a potential quite impactful proposal? And then secondly, in recent months, we've had a flurry of announcements from the AI native players in Legal, such as Harvey and Legora based on their communication, they're rapidly scaling with ARRs in excess of $400 million combined and some very strong user growth. You talked about healthy adoption of your own workflow products. Could you give us any kind of comparable metrics on engagement or revenue to help us assess market shares?

I'll ask Nick to cover the first one, and then I'll come back and talk about the second.

Yes. Will, obviously, what you're seeing in our numbers today around science demonstrates the pace of growth in the scientific research world, the pace at which scientists are operating, the pace at which they're making discoveries, the pace at which they want to be published. And they're all using new tools and new efficiencies to increase the pace at which they're operating. And that's why you see the very strong growth in submission volumes and that, of course, is coming through into strong growth in published volumes. What government policy does in individual countries around the world, that's been changing all the time. And you could have a conversation going back for years about different things that have changed. But the fundamental drivers on a global basis remain the same. And that's what's ultimately going to drive the business going forward.

Yes. On legal AI, I think it's important to look at this from a broad perspective that the legal information space in the U.S. is order of magnitude a $5 billion market where we operate. We're information based, and we are information-based analytics and decision tools, which we increase the value by AI enabling them and adding process and decision tools and workflow tools on top of the verified trusted content. And we put all the analytics, the AI tools that come from the sort of frontier lab companies, if you want to call them that, we include those inside our products, inside our secure environment with a verified content, et cetera, et cetera. We are an information-based company with higher value-add AI-enabled tools on top. That's what we sell, and that's our value add.

The legal tech software space is order of magnitude, $25 billion today, depending on which analysts you look at. Many people believe that, that's going to grow to $75 billion to $100 billion over the next 10 years. That's process enabling in the legal workflow, in legal companies, as well as in financial institutions, in corporations, in private equity firms and so on, where they, today, an average large company might have 100 different software technology providers. AI can improve those processes, add significant value to the whole legal industry, which is order of magnitude, maybe, call it, $500 billion in revenue.

And when you can provide new technology tools to improve the efficiency of $500 billion industry, you're not just going to have the old 1,000 software providers, of which an average player uses 100 and 200, you're going to have several new ones, several hundred new ones competing for that space. That just illustrates to me how big the opportunity is to help that industry. We are not going to be the leader in the $100 billion workflow and software industry, but we are going to be able to contribute value to our customers in content-related workflow tools, then we can see that from our customers. We have, as you know, hundreds of thousands, many hundreds of thousands of users, many tens of thousands of institutional customers paying corporate customers for our products.

We have very high usage of those tools, and we are in the technology-enabled higher value-add information-based tools. The fact that there are other players out there doing well to me is a confirmation that the space is important and there's an opportunity to add value. We will capture a portion of that. But we are not attempting to or planning to go and compete directly against tech or software players in the workflow space.

Just to come back on the OMD proposals around journal subscription costs and APC charges. Is it fair to understand from your comments that you wouldn't expect any meaningful impact in the event of the introduction of those reforms?

Yes. We make sure we provide our customers with the choice so they can publish in the journal that meets whatever criteria they need to follow. We've been doing that for a long time. I see no reason why we can't just -- be able to continue to do that.

The next question comes from Thymen Rundberg of ING.

Three from my side. Yes, this year, you've completed around 80% of the announced buyback program by -- in the first half year compared with roughly 2/3 to 70% in recent years. So stepping back from timing of execution, does the larger and faster deployment of capital into buybacks reflect a different assessment of the relative attractiveness of buybacks versus other uses of capital, whether that's M&A or other organic investment opportunities? I know you mentioned that organic keeps being the #1 priority. Or should we continue to view -- yes, should we just continue to view capital allocation as unchanged despite the change in pace and size.

And then on Risk. So Risk delivers 8% organic growth for a number of years already despite its scale. When you look at the different subdivisions, what do you think ultimately determines the sustainable growth rate of that business? And where do you see the greatest scope to outperform those assumptions over time? And then lastly, just a quick one on the Doctrine acquisition that you proposed in April. Could you just elaborate a bit on what you think that brings -- or would have been difficult or time consuming to build internally and how you think about the broader M&A space in Legal AI, particularly?

I'm going to ask Nick to cover the first one. I'll talk about the second, and then we'll get Nick back on the acquisition again.

Yes. On the buyback, absolutely no change in approach to capital allocation. As you rightly say, the primary focus is on supporting the business and in particular, on the organic development. So obviously, all the capital investment, all the CapEx required internally is the #1 priority, but this is a relatively capital-light business. After that, we do all the M&A that we see works can enhance and accelerate that. It clearly varies from year-to-year depending on what comes up. we haven't increased. And then we -- only after that, that we can think about returns to shareholders, obviously a growing dividend, that's about half the earnings.

And the buyback tends to be the balancing figure, if you like, to keep the leverage in the right sort of place. Clearly, we're mindful of where the share price is relative to a year or 2 ago. And certainly, in making the decision on the buyback in February as we did increasing it, knowing that would likely push us up in the leverage range, we're using that leverage range of 2 to 2.5x in a thoughtful way. But an overall approach, still following what we were doing before.

On Risk, as you -- I'm sure you know that over the last decade, we pretty much average an 8% annual growth there. And the range has basically been 7% to 9% if you exclude the first few months of COVID when the world stalled a little bit. So we always try in every single subunit there to capture upside. The main driver of the strong growth of that ongoing 8% growth in risk is the continued development and rollout of higher value-add decision tools. And we keep launching those, testing them, selling them and rolling it out. Typically it takes about 5 years to fully roll out a product across that industry.

We keep doing that, and we keep trying in each subunit to grow faster, to roll them out faster. You could argue that the technology we're seeing today -- with the technology we're seeing today that we could develop them and launch them faster, but the industries we serve still operate at roughly that same introduction, installation and rollout pace because they're complicated industries, they're regulated and they operate at a certain pace. We would love to be able to grow faster in any one of our units, and sometimes we do. But it's now almost a $5 billion revenue division. And in order for this to permanently grow faster, we would have to have several of the units pick up on the same schedule for a period of time. We believe it's quite possible to do that. We want to do that, but I don't think it's something you should count on or build into your projections.

Yes. And your question about the Doctrine acquisition, it's a very good example of the sort of thing that we're looking to do to enhance and accelerate the organic development. Doctrine has been around for about a decade building a very valuable content set in France. They've developed some very good workflow tools and some AI around those specific content sets. And we see how that will fit -- complement what we're doing in France very well. And we are -- we think we're the natural owner of that business, and that's exactly the sort of thing that we would look at. Of course, in the overall group sense, it's quite small. So -- but it gives is a good example of the sort of thing we're interested in to enhance and accelerate the organic development.

Our next question comes from Steve Liechti of Deutsche Numis.

I've got 3 quick ones. One, just on Claude for Science, any particular comments there in terms of what uptake of LeapSpace has been or feedback from customers in terms of using Claude for Science relative to LeapSpace? Second, just remind us on Arabian travel, how big the international -- I'm really kind of thinking Western visitor or exhibitor numbers are for that particular show in percentage terms, that would be great. And then just on the travel like-for-like at 6%, which is below previous years. You referenced travel disruption. Do you mean specifically in the Middle East overall? So is that 6% being affected by something else or specifically the Middle East?

I'm going to ask Nick again to cover the second, but let me first make a comment on Claude for Science. The way we look at this is these are tools that are being provided by others. There will be many of these kinds of tools and announcements coming out. There will be lots of providers of these tools. The way we see it is at the moment, these are the LLM providers using additional workflow layers to their models. And they're mostly designed to support the conduct of the science itself often in drug development. They're going to continue to do that. And we think that's a very good thing for science. Anything that helps improve the productivity of sciences of scientists and or the productivity of spend on science is a positive driver of our business.

We're focused on the research, the publication of science and the use on research around published science and the research workflow around that LeapSpace support research workflow, idea generation, literature exploration, collaboration, sourcing, funding analysis, drafting, comparisons inside a completely verified trusted, secure private confidential space that is about your research. And we see Claude for Science and many of these other tools that will come out as a positive enablers of the pace of scientific discovery and very positive contributors to the industry and therefore, an additional driver to what we do, not a competition at all.

And Steve, your question -- second question about Arabian travel, just as a reminder, as I said earlier, we're talking about the shows that are yet to run in the region for Exhibitions about 3% of divisional revenues or less than 0.5% of group revenue. You're asking about one particular show within that. It is a heavily international show, but -- and exactly how that plays out, we'll see, but I just keep the scale of it in certainly relative to the group as a whole, keep that in mind. I think your other question was about the 6% growth in Exhibitions in the first half and our reference to travel disruption. That was really referring to travel from and through the Middle East to our other events, to events outside the region.

We have obviously a number of quite significantly international events elsewhere in the world and the participants from the Middle East or people traveling through the Middle East, and that had some effect, which is why the ongoing growth in Exhibitions of 7% to 8% that we've been at for a little while. That's why it was just moderated a little from that in the first half.

[Operator Instructions] Our next question comes from Christophe Cherblanc of Bernstein.

So [indiscernible] from my side. The first one is on STM. What should we expect from the - what should we expect from the settlement made by Anthropic with publishers? Because it seems to me it could be almost 1% of STM revenues. So is it going to show up as revenue and operating profit at the same time? Will it fall in H2 this year, '27? Anything on this would be super helpful. And the second one is on Print. The resilience of profit was very good in H1. Is that the pattern we should expect for the full year '26?

I'm going to ask Nick to cover both of those.

Yes. Christophe, I'm sure you'll appreciate I won't comment on individual legal settlements and the like. But what I would say is any receipt from a legal settlement is not revenue. It's an offset within costs and as is the expense of defending these things, so whatever, not revenue. And your second question on Print. Yes, I think as we've said, our objective is to reduce our exposure over time to the remaining print activity. You saw that again in the first half with another step down in revenue. But we're doing that through outsourcing, through joint ventures and things. So as well as the natural decline, you get some faster decline of that. But we're looking to hold on to as much of the profit as we can as we do that. So this sort of high single-digit decline in profit is a sort of objective and the territory we've been in and look to continue to be in going forward.

Our next question comes from Jo Barnet-Lamb of UBS.

It's Jo from UBS. I think just 2 left. Firstly, both Legal and STM ticked over to 6% and 10%, respectively. You tweaked STM outlook up to continued strong, but you didn't alter the Legal outlook language from continued strong. Can you just talk briefly on how you view the sustainability of the acceleration in Legal? And sorry to be pedantic, but does strong stop at 10%? Or could 11% in theory, also be turned strong? And then secondly, in Insurance, as auto risk evolves with advanced safety features and autonomous driving, should we think of this as a shift in the type of data required rather than a reduction in data demand? And then as you -- are you evolving your data sets beyond sort of historic areas of strength such as driver characteristics towards vehicle usage, location and otherwise? Some thoughts there would be great.

I'll ask Nick to cover the first, and I'll come back and talk about the second.

So Jo, in our language, 6 and above is strong, and that's the terminology we use for everything from there. So you shouldn't read anything into the way those words are put together. It's just -- that's our language and that just reflects the performance of the business.

But I think you just have to combine that with the outlook. The statements we have on our objectives. We've said and keep repeating that our objective in both Legal and STM is to continue on the improving growth trajectories in both of them. But we have had a tick up on growth rate pretty frequently in Legal recently. And I want to point out that 85% of the Legal base is subscription and 80% or so is subscription in STM. So that most of the time, you should expect continued improvement in growth rate to come through gradually and not continue to tick up every 6 or 12 months, we may have achieved it for a while now in Legal. But our objective is to continue on these improving growth trajectories in both of these divisions. And as you can hear from us probably, we do think that we're just at the beginning of capturing the value upside of AI-enabled tools in our product suite to our customers over the next several years.

When it comes to autonomous driving, the way we look at the whole industry, including insurers, car manufacturers, technology companies will continue to evolve, use different data sets, new technologies and get safer as a principle, just like it has historically with different safety features, electronic tools, EVs, et cetera. But what we've seen when that happens is that the accident frequency often drops a bit, but the severity -- cost of repairs goes up significantly. And in that dynamic, the environment gets significantly more complex. And that increasing complexity and changing environment creates opportunities for us to add more value, just like it has in the past. And we believe that if you get more full self-driving coming through and you transition from entirely manual to partial autonomy at some point towards full autonomy in part of the vehicle park on the roads, we're going to see decades of this increasing complexity and the interactions between different cars and different drivers between manual and partially or fully self-driving that creates a complex environment where you have to adjust as an insurance company, your pricing to the new risk assessments.

And the faster it keeps going, the bigger the opportunity is for a third party like us that keeps collecting data sets from different sources and new data sets to then help them navigate that. We have seen this already over the last few years. As you said, we had some traditional data sets many years ago, but we've seen a significant increase in the rate of new data sets being available as these tools have come out over the last few years, and we've collected significant data sets also from car manufacturers on the software, on the technology tools and how they interact and the probability of different types of risk profiles in that.

So -- the data sets are increasing significantly. The sophistication of our model increases as the complexity of the vehicle park and their interaction increases, bigger opportunity for us to contribute for a very long period to come.

Ladies and gentlemen, with no further questions in the question queue, we have reached the end of the question-and-answer session. I will now hand back to the CEO, Erik Engstrom, for closing remarks.

Thank you all for joining us today. Enjoyed talking to you, and I look forward to talking to you again soon.

Good morning, good afternoon, and welcome. I'm Rick Trainor, the CEO of Business Services within LexisNexis Risk Solutions. I've been with RELX for over 20 years, having joined through the 2004 acquisition of Seisint and have been leading business services since 2009.

Today, we're going to give you an update on business services with a particular focus on Fraud & Identity. I'll walk you through our strategy, highlighting the sophisticated data analytics capabilities. You'll then hear about our technology approach from Risk CTO, Vijay Raghavan.

We'll then bring the Fraud & Identity business to life with 2 deep dives. Kim Sutherland, VP and Global Head of Fraud & Identity will take us through a customer case study on how customers use our Fraud & Identity solutions.

We'll then have Matt Adams CTO and Co-Founder of IDVerse; and Dan Aiello, Chief Product Officer and Co-Founder of IDVerse, walk through a new account opening case study. Afterwards, Vijay and I will come back for Q&A.

Let me start with where risk fits within RELX. In 2025, Risk represented around 36% of RELX' revenue and about 39% of the profit. Our full year revenue in 2025 was GBP 3.5 billion or about USD 4.6 billion. The long-term fundamentals of the risk business are strong. We have continued on a strong and consistent growth trajectory with average underlying revenue growth of 8%. While there will always be some fluctuations across cycles, you can see a consistent 7% to 9% growth over the last decade with the exception of 2020 due to COVID.

Our main business segments have an average growth rate roughly in line with the divisional average and fluctuations tend to cancel each other out over time. We're not only a high revenue growth business, but also high margin. And a key factor in this is our scale and the ability to reuse our unique data assets, our technology, and linking and AI capabilities in each of our key segments. This coupled with a focus on continuous process innovation helps us manage cost growth below revenue growth. And the gap between underlying revenue growth and underlying adjusted operating profit growth is widening with the advancement of technology.

While we operate in global industries with structural growth drivers, innovation is a key reason for our consistent performance. We continue to enhance the value we deliver to our customers through new solutions and capabilities that are deeply embedded into their workflows.

In this slide, the orange part of the bar is growth that comes from new products. We define those as products launched in the last 5 years, which is a typical adoption cycle to roll out new products for us. And new technology is helping us develop and launch products at a faster pace.

We have 4 key capabilities that we leverage to drive innovation and add more value to our customers. First is our deep customer understanding. We work in close partnership with our customers to help solve some of their most fundamental business challenges. Our solutions are deeply integrated into our customers' workflows, where we help inform or automate key decisions with over 90% of our transactions being machine to machine. Our deep understanding of our customers' businesses combined with our core skill and innovation is a key driver of our success.

Our second key capability is our leading data sets. This is the foundation of our business. Our data assets have been created over decades of licensing, aggregating, linking and building data. As for scale, breadth and depth of this data, we have tens of billions of public records and data elements across tens of thousands of sources. Most importantly, we have contributory and proprietary data sets that are unique to us and are a core part of our differentiated value proposition.

We now have over 25 contributory databases across risk. This is where our customers contribute their data to us so that we can provide them back risk analytics across the market or across industries to solve specific use cases.

I'll speak to this in the context of the business services use cases shortly. We continue to grow the depth and breadth of our data, and we're also adding different types of data to provide greater risk insights to our customers. It's important to note that we serve customers that operate in highly regulated markets. It's incredibly important that the answers that we deliver to our customers are highly precise, accurate, explainable and compliant.

Our third capability is our advanced linking and analytics. We have a long history of using AI and other advanced analytical approaches. This underpins our linking capability, which allows us to connect these vast amounts of disparate data points to create one unique view of an individual or a business. We also utilize our sophisticated analytics and AI in the solutions we provide to our customers through scoring models, attributes and diagnostic tools, which enables them to make decisions.

Our models have been refined and improved over decades, utilizing important customer feedback loops. We continue to apply the most sophisticated approaches to ensure our products provide our customers with industry-leading quality and accuracy.

And finally, our technology platforms. We have a fast, scalable platform that allows us to ingest more and more data and seamlessly plug in new AI technologies. This also allows customers to connect to our solutions seamlessly. We serve 4 business segments within the Risk division, where we help our customers assess and manage risk and identity fraud.

Business Services is the largest segment and represents nearly 45% of Risk revenue. This is where we'll focus today. Insurance is the second largest segment and nearly 40% of revenue. Specialized Data Services is just over 10%, and Government is about 5%.

Now let me walk through business services in more detail. This slide shows you business services revenue by geography, solution and type. On the left, you'll see that we now generate nearly 30% of our revenue from outside of the U.S. Since I last presented, we've become more global, and non-U.S. expansion will continue to be an important growth opportunity and is being driven by the mix of solutions we offer and the relative maturity of each market.

In the middle chart, I'm highlighting the proportion of revenue from local solutions those built on local data assets for local markets and our global solutions, which are those data solutions applicable worldwide.

We continue to expand our portfolio of global solutions, which now represent over 45% of our revenue. We have a long runway with these solutions in both the U.S. and non-U.S. markets, and we expect this to continue to be a core part of our growth engine going forward.

Finally, on the right chart, you can see we have a balanced mix of revenue from subscriptions and transactional solutions and our transactional revenues are under long-term contracts with a volumetric component. There are very few onetime transactions. We serve a large, diverse customer base with over 18,000 customers in more than 180 countries and territories. Our solutions are developed to meet the needs of customers of every size from the world's largest and most sophisticated businesses, to small and midsized businesses.

Our revenue concentration is quite low with our top 30 customers making up less than 30% of our overall revenue. Our solutions are used across industries, including financial services, which was our largest customer segment, digital service providers like telcos and retail and e-commerce and a long tail of others. We also provide solutions that other parts of the risk division take to market in the insurance, government and health care sectors.

We help our customers assess risks associated with the consumer or business or a transaction, whether that is fraud, compliance or credit risk. This helps our customers make higher confidence decisions makes the transaction process more efficient and safer for consumers. Our business is segmented into 3 primary business areas. Fraud & Identity solutions account for a little more than 1/3 of the revenue and is the largest part of our business. It will be the primary focus of today's discussion.

What we do here is help our customers evaluate if an identity exists, can it be trusted and whether a transaction is legitimate. We do this by analyzing hundreds of digital physical and behavioral attributes associated with an identity in the transaction to help our customers understand which they should allow through these systems without friction in which are higher risk, requiring additional levels of diligence, whether a particular transaction should be rejected outright.

Financial Crime and Compliance accounts for a little under 1/3 of our revenue and is our second largest segment. In this segment, we deliver a suite of solutions that help our customers comply with global regulations such as know your customer, anti-money laundering, counterterrorist financing and any bribery and corruption statutes. We do this by validating that the identity exists and the identity attributes are accurate.

We then screen the identity details against various watchlists such as governmental sanction list, economic sanctions and politically exposed individuals. And although an identity may have been determined to exist, our customers must also demonstrate that it is legally permissible to do business with them or to treat them with a higher level of risk.

Finally, the balance of revenue comes from credit business and other risk solutions. Here, we provide a range of specialized solutions, including alternative data solutions for understanding the credit worthiness of consumers and businesses, along with due diligence tools. All of our solutions are underpinned by a combination of highly differentiated data assets and complex analytics, which I'll talk about more in just a moment.

The challenges facing our customers are large and global and only getting bigger and more complex. The number of fraud attacks and associated fraud losses are growing, driven by automated bot attacks and AI fraud schemes. There are more sanctions and regulations being imposed that must be met by an increasing number of organizations. Cross-border transactions, cryptocurrencies and other new transaction methods make tracking money flows and compliance harder. And consumers are increasingly using nontraditional borrowing types like Buy Now, Pay Later. That, coupled with changes to traditional credit reporting, makes traditional credit files less representative of risks supporting the need for more alternative credit data solutions.

With the rapid evolution of AI, bad actors are operating faster and at a larger scale than ever before. There are more sophisticated deep fakes and synthetic identities and evolving fraud schemes in more systemic attacks. And for our customers, this means that serving their customers and growing their business is harder. It's increasingly difficult for them to assess risk and establish trust during a transaction flow resulting in outsized financial and operational impacts.

We are incredibly well positioned to help our customers solve these growing challenges. We layer intelligence at every point as our customers interact with their customers, enabling our customers to see a full picture of risk associated with all aspects of a consumer interaction. Our solutions are deeply integrated into our customers' workflows, and most of our solutions are machine to machine, meaning within a fraction of a second, as the customer is interacting with their customers, we can assess that this is a legitimate person or an agent that they want to do business with or operating on a trusted device with identity attributes and behaviors that are consistent with recent patterns.

At each stage of the process, we verify the connection between the consumer, the device, the agent and provide intelligence around the risk that helps them make higher confidence decisions. This makes the transaction process smoother, more efficient and safer for consumers. As a result, our customers can grow confidently, onboard and protect legitimate customers without friction and operate more efficiently and in compliance with worldwide financial regulation.

The way we do this is by providing our customers with a comprehensive and multidimensional view of a consumer or a business, including attributes tied to their physical identity, their digital identity and their behaviors. This helps identify when there are patterns during a transaction flow that appear unusual and potentially risky. Our solutions enable our customers to confidently assess whether they should trust the person, the agent, the device and the behavior associated with whom they're transacting. This deep view of our consumer or business is what fuels our analytics engine. The scale, breadth and depth of our data assets are truly differentiated.

I'd like to draw your attention to the following stats that help demonstrate the scale of our network. We cover virtually all the adults in the United States. We process over 1 trillion sanctions annually. We processed roughly 145 billion digital transactions annually. That includes 81 billion logins, 2 billion new account creations and 28 billion payments.

There are 4 primary sources of the data in our risk intelligence network, including 2 foundational and 2 proprietary sources. The first foundational source is our public records repository. We have tens of billions of public records from tens of thousands of sources that we've built over decades. Some of the data is no longer publicly available, and some is theoretically public but extremely difficult and complicated to collect because of the format, general data source availability or requires manual collection.

The second foundational source is our license data, which comes from thousands of different sources to add further intelligence, breadth and context. For these sources, the usage is commercially controlled and regulated, meaning we can only use them in certain ways in our solutions. And then we have our proprietary network driven sources.

First, we have our contributory data assets, which are built through customer interactions with our solutions. To benefit from the value of these solutions, customers must contribute their activity to the risk intelligence network. So each time a customer transaction happens, the input data, the data attributes, the patterns of behavior and the outcome of that transaction is captured. This means our data asset is becoming richer and deeper with every transaction.

Finally, we build proprietary derived attributes which further enhance an identity profile or its correlation with risk. All 4 dimensions are combined to create a longitudinal network of risk insight that grows over time. This vast data alone has little value to our customers. We transformed this data utilizing sophisticated analytics and AI into specific signals and scores and feed that into our customers' workflows to assess risk in real time.

We have created a great virtuous cycle. As we process more transactions and outcomes, we are able to see more signals about risk and how patterns of risk are evolving, which allows us to create even stronger signals of risk which makes our products stronger and delivers more value to our customers.

Today, we see over 400 million transactions every day, and the number continues to grow as we add new customers as our existing customers grow their usage and our customers deepen their relationship with us. Our solutions deliver better outcomes for our customers and create significant differentiation in the measurable value uplift we provide. While the solutions we provide, our customers represent a small part of their cost base. They have significant positive impacts on the economics of the overall business.

We identify and stop more fraudulent transactions even in the hardest to assess bands. We deliver less false positives, allowing more good customers through without friction. We make sure only the highest risk transactions are routed to high-cost methods of review. The net of this is higher revenue and lower operating costs for our customers.

Customers deepen their relationship with us over time. The left side of this slide shows an example of our relationship with the U.S. financial institution. We initially sold this customer an identity verification solution to improve their KYC program. As the customer recognize the value we provided, they adopted more solutions, adding new capabilities across more use cases, such as account management and fraud prevention.

While every customer is different, the shape of the journey is very similar across most of our customer base, we price to capture a small portion of the value we provide. As customers see strong price to value of our solutions, they increase the number of products they purchase and the depth of our integration into their workflow. As we continue to innovate, we expect that all customers will continue to layer in more capabilities and expand their relationship with us in this way.

This slide highlights our strong track record of growth over the past 25 years. We have expanded primarily through organic innovation, supplemented by targeted and highly complementary acquisitions. We have a very disciplined approach to M&A, evaluating hundreds of new technologies, new solutions every year. Many of our successful acquisitions like ThreatMetrix and Emailage, started as commercial partnerships which gave us a deep understanding who has leading capability and what the combined value proposition is for customers.

IDVerse is our latest acquisition, completed in February of 2025, which added AI-powered document authentication and deep fake analytics to our portfolio. Before we acquired IDVerse, we assessed nearly every provider of scale in the space, either through partnership or other commercial discussions, comparing technologies, comparing analytics, and testing their ability to catch fraud. This gave us confidence that we're acquiring the most sophisticated capability in the market.

You'll see Matt and Dan demo this in just a little bit.

Our customers' challenges are far from static and our evolving solutions, robust data network and ongoing innovation keep us well positioned for future success. And now let me turn it over to Vijay to take you deeper into our analytics and technology approach.

Thank you, Rick. I'm Vijay Raghavan, I'm the Chief Technology Officer at Risk. I've been in this role for almost 15 years, and I've been at the RELX for almost 25 years. I'm also the Chair of the RELX Technology Forum, which has best practices across the RELX divisions. Rick briefly touched on our core capabilities, and I'd like to walk you through our analytics and technology approach in more detail.

Let's start with a function of technology at RELX. At a fundamental level, we are the enablers of the innovation engine you have heard so much about today. We help our businesses execute against our growth plans by investing in the right technology capabilities to enable our teams to innovate quickly and efficiently with the right tools and to ensure that our systems are flexible, reliable and scalable.

Given the nature of our business, it is the role of technology to make sure that we have highly secure environments to protect our customers' data and IP and to adapt to changing regulatory requirements. An integral part of technology's role is to continuously automate and optimize through the improvement of our processes and our tools.

Technology is a real source of competitive advantage across RELX. At the heart of that is our people and the ability to stay at the forefront of the evolving technology landscape. These are highly innovative teams with deep experience and expertise in data analytics and AI and ML techniques and who are motivated to use technology to improve outcomes for our customers and for ourselves.

We have a long history of using advanced technology within Risk. We first created our big data technology in the 1990s, long before big data was a buzzword. We then created a proprietary machine learning-based linking technology in the mid-2000s. We first started talking to you about big data and usage of analytical algorithms back in 2011. By around 2015, we had been using AI and ML techniques for over a decade, and that's when we first started sharing externally about how we have evolved in AI and ML tools and processes into the fabric of our data and our technology.

In 2018, we talked to you about how we use supervised and unsupervised learning in our AI solutions and how we've assessed and evaluated multiple algorithms to provide the greatest value to our customers. In 2023, I spoke about how we were integrating real-time machine-generated data into an existing fabric of public records data, contributor databases, device intelligence and digital identities to give our customers even more comprehensive solutions.

I also spoke to you then about generative AI and how it will give us greater scale to innovate, for example, around knowledge extraction from our data repositories and automated code generation. All these presentations remain available on our website. What we are doing today with our technology is consistent with our history. We are constantly evaluating new tools to evolve our approach to support better, faster, cheaper innovation, and we use the best and most appropriate tools for the job at hand to create even more compelling products.

The way this has evolved since the last time I spoke to you in 2023 is that we are embedding generative AI and agentic AI tooling into our technology stack. However, since some of these AI tools come with side effects, we have built a trusted AI infrastructure around these tools in order to not compromise the quality and integrity of our solutions. It is paramount that we continue to offer our customers a trusted, reliable compliance solutions they have come to expect from Risk even as we adopt new AI techniques.

Here's what I mean by that. This slide describes the layers of our technology stack. Across these layers, we use a variety of technologies, including open source, third-party and proprietary solutions. At the bottom of the stack is our infrastructure layer where we use third-party cloud tooling such as servers, networks, storage, databases and other infrastructure as a metered utility. These tools are broadly available in the market and are not unique to us.

What is important is how we deploy these tools, which we do in a cost-effective and flexible manner. The more important things in the stack are in the middle and top layers. And in the middle is an abstraction layer, which is a proprietary approach that gives us much greater control and flexibility over how we use third-party services. We can now easily leverage emerging third-party innovations, including generative AI, LLMs and Agentic solutions offered by cloud vendors and hyperscalers and switch between these third-party services easily.

For example, we can integrate a new LLM into our platform within a few hours but what's just as important is that this abstraction layer helps ensure that even as we adopt new AI techniques, we don't make trade-offs between the speed and quality of our answers or between the quality and consistency of answers or between the speed and transparency of answers. That is critical for our customers.

Resting above the abstraction layer is the applications and product layer that represents our core IP, where we ingest data at high speeds, link the data with great accuracy, boil the data down to discrete elements that we call entities and then build solutions that are easily consumable by our customers. We have been doing this for a long time, but we are continually finding ways to make these approaches better.

On this slide, I'd like to drill down just a little deeper into the abstraction layer and the applications and product layer. Let's start with the abstraction layer. A key element of this layer is our trusted AI infrastructure that you see depicted on the bottom right. This is crucial because as we deploy new AI solutions, including agentic AI and generative AI, we can assure our customers and regulators that the decisions made by our solutions are transparent and defensible. That's what I mean by not having to make trade-offs.

Our trusted AI infrastructure within our abstraction layer provides us with AI validation guardrails that are superior to pure-play LLM-based solutions that claim to be nearly as good, but are either written with bias, which is unacceptable to customers, or they are opaque and nondeterministic which is unacceptable to regulators.

The top of the slide shows 2 examples of our applications and product layer. Omega AI is a modernized data fabrication process. This replaces our previous generation of technology with one that is cloud-native and AI enabled. One of the big advantages of our new approach is that we can ingest data incrementally with near real-time propagation of data into our products, which significantly improves the value we provide to our customers.

The entity database is the other example. It serves as another force multiplier for us because it is a canonical entity-centric representation of data that gives all our products a shared model of entities and relationships. Entities could be people or businesses or vehicles or drivers licenses and so on. Essentially, we are using AI to create consistent, reusable and well-connected entities across our platform for us to not only be able to build our products more easily, but also to create an ontology of digital entities that adds much greater value to our customers.

Now let me touch on how we build our products with this technology stack I just described. You have seen this slide before, but it is an incredibly important one. It demonstrates how we get from data to specific actionable insights that help our customers make decisions. Rick touched on the scale and breadth of our data assets, but Big Data itself is not of much value to our customers. Our technology transforms big data into small actionable intelligence at scale and at high speed to add value to our customers' decisions, for example, in the form of identity authentication or device authentication or agent authentication.

And to give you a sense of scale and speed, our ThreatMetrix product verifies 200 different data points for each transaction it sees within sub-seconds and it does this across 400 million transactions per day. We do this by layering advanced analytics on top of our data and to cluster, link and identify patterns to improve our solutions. This is what we call Extractive AI and it is at the heart of Risk's competitive advantage.

As Rick said earlier, over 90% of our transactions come from machine to machine in the form of scores or attributes as opposed to generated text that a customer needs to analyze or interpret. So it is incredibly important that the answers that we deliver to our customers are highly accurate and compliant and that they are consistent, meaning a customer always receives the same answer in the same situation with the same inputs. That is quite difficult in the probabilistic way LLMs operate, where answers may evolve over time.

Our reliable, deterministic approach is critical because of how we integrate with customers and also because of the regulatory nature of our customers' use cases.

We have honed our proprietary algorithms over decades to continue to improve these parameters and to create more and more sophisticated techniques, which continue to enable industry-leading accuracy with fast cycle times and at lower costs. Due to the nature of the Risk business, extractive AI is fundamental to our solutions. However, we also deploy generative, agentic and other AI capabilities. We layer them on top for extractive AI approach.

You'll see a few examples here on this slide. The first example is related to image analytics and insurance. In 2023, we talked about a Flyreel product, which is an AI-based image capture solution that allows a layperson to capture a video of their property in home insurance underwriting context, or a video of the automobile after an accident in a claims context. We have continued to find ways to improve the way videos are automatically analyzed in the background using increasingly advanced algorithms to process those images, assess context and risk and extract relevant data into our platforms.

The next example is LexID, which is our proprietary approach to link our data assets together in a highly accurate manner. Our linking underpins nearly all of our solutions and we have continued to refine our algorithms to improve our linking over decades. Today, we have industry-leading linking accuracy. However, we continue to find ways to improve that linking to get even closer to perfection because every incremental bit of improvement in accuracy improves value for our customers. We are now using generative AI and agentic AI techniques coupled with a human-in-the-loop to map raw unstructured data into structured data even more accurately and to further improve our linking accuracy.

The third example on this slide has to do with the IDVerse solution, which you will see in a demo a little later in the presentation. Fraudsters are getting more creative every day with deep fakes. Our proprietary neural network within the IDVerse platform has been constantly enhanced to detect new fraud patterns. For example, 2 years ago, it would have been sufficient to rely on facial landmarks, eye movement and lip sync to detect fraud. Fraudsters are now able to get past that.

So now we've enhanced our neural network to detect liveness by using skin spectral analysis and optical flow analysis which tracks involuntary movement of facial muscles due to blood flow.

Our neural network handles document authentication, biometric face matching, liveness checking, depth-based 3D analysis, injection attack detection and deep fake classification all in a single pipeline, which makes it incredibly sophisticated in identifying fraud.

Again, you'll hear more about IDVerse a little later. These are just some examples of how we are continuing to leverage more and more sophisticated technology to improve the value we deliver to customers, and there are many more.

We are also using our technology capabilities internally to enable us to improve processes and make our people more effective in their day-to-day work. You will see a few examples on this slide. I won't walk through all of these, but I will touch on a couple of examples. In our technology function, we are actively employing AI-assisted coding, which is clearly helpful for product development.

But especially interesting is the value that it adds to the upgrading of systems, implementing new technology and advancing our cybersecurity defenses. We do so aggressively, but judiciously in keeping with the approach of including a human in the loop. There are many more examples across all of our functional areas, some of which you see here.

We continue to find ways to apply technology internally to operate with more agility and more effectively, which in turn allows us to innovate faster and serve our customers better.

So to wrap up, I hope you walk away with a better understanding of how we use technology at Risk. Our technology and analytics approaches play a central role in enabling rapid, agile and low-cost innovation across the business. We have used AI for decades and continue to deploy the most advanced methods to constantly refine the accuracy and value of our products and enhance the effectiveness and efficiency of our internal teams. And as technology continues to get more and more advanced, we are well positioned to adopt these tools quickly, deploy them in the most appropriate manner and strengthen our position over the long term.

With that, I'll turn it over to Kim Sutherland to bring our technology to life with a customer case study.

Thank you, Vijay. My name is Kim Sutherland, and I'm the Global Head of Fraud & Identity. I've been with the Risk for 20 years, and during most of that time, I've been focused on building our commercial market strategy for our portfolio of Fraud & Identity solutions.

The way that consumers interact with business is evolving and increasing in complexity. During a single interaction, a consumer may log into an account on their phone, move from a mobile app to a website, issue a real-time payment and initiate an account-to-account transfer. We are seeing a growing number of interactions through more devices and channels from mobile browser and digital wallets and now the emergence of agentic commerce, and this growth means more opportunity for fraud.

Recognizing trusted behavior and detecting anomalies in real time across every device and every channel is no longer optional. It's a baseline expectation. Vulnerability to fraud attacks persists across the entire consumer life cycle. We help customers reduce fraud by layering defenses at each of those touch points.

The first layer, digital and identity assessment, uses device, location, behavioral, bot and agent intelligence, to establish trust from the very first signal, in addition to basics like identity attributes, such as e-mail address, name and phone number, they're also verifying.

The second layer applies decision analytics. Adaptive fraud analytic models, machine learning and orchestration to identify anomalies and velocity patterns in real time.

The third layer adds authentication. From passive methods to bind a trusted device to active methods, including biometrics and document authentication. In the fourth layer, investigation and review closes the loop with forensics, case management and even the incorporation of fraud feedback. We leverage an integrated platform for dynamic and coordinated use of these solutions and underpinning the layers is our risk intelligence network, ensuring that every signal adds the required context to assess risk for every interaction.

How these capabilities are deployed is determined by the customer. This enables a fast, frictionless experience for the vast majority of consumers and transactions that are low risk, while providing strong protection when there are signals of fraud. A consumer transaction can seem very simple. But behind that moment, thousands of data signals are being collected and analyzed. Fully automated risk decisions are running in real time and fraud risk models are scoring the interaction. This is done in approximately 85 milliseconds and over 400 million times a day.

And at the core of that decision engine are 3 fundamental questions. First, who is this? Does this identity device behavior and combination of signals had any history within our network. Identity recognition is the foundation of trust. Second, can they be trusted? Are the attributes accurate? Are there any suspicious activities associated with this behavior, this device or this identity and critically, is this person a victim themselves, potentially being manipulated without even knowing it.

Third, do we need more proof? Ambiguous signals require further verification. Our approach turns disconnected signals into a single connected digital identity. Every digital identity creates data in our network, an e-mail address, a device and how you interact with it, a phone number, a billing address, a payment card, a location.

In isolation, each of these signals tells a partial story. But connected together, they can reveal something far more powerful, a trusted identity. A typical user has 1 to 2 e-mail addresses, 2 to 4 devices and 2 to 4 payment cards. When something shifts a new device, an unfamiliar location, automated filling of identity attributes or behavioral signals that suggest the users being coerced, which is a hallmark of sophisticated fraud. Our network recognizes those moments.

Let's walk through an example of a consumer logging into their account. What I'm showing here is an example of how our customers protect digital logins while keeping the experience seamless for trusted users. On the left, this is the consumer experience, a familiar log-in screen and on the right is ThreatMetrix and BehavioSec working together in real time. As soon as a user lands on the page, we began building risk context using device, network and hardware intelligence.

Now I'll complete the log in, and you'll see the outcome is a pass. And near real time, tens of milliseconds, all of these signals are evaluated behind the scenes with no impact to the user experience.

What's important for our customers isn't just the decision. It's understanding why the decision was made and that's where reason codes come in. Reason codes provide transparent, explainable insight into what contributed to trusting this user. In this case, we're seeing multiple positive signals come together. This is a recognized user logging in from a known device. There's an established historical behavior over time, not a one-off interaction. And this identity is trusted across the digital network. Together, these signals create high confidence that this is a legitimate low-risk user.

Now let's look at a different example. Here is a consumer registering an account with one of our customers. This is the customer's first time seeing this consumer and the only data they have is an e-mail address in the device, on its own that's not enough to confidently assess risk. With our network, we layer in significantly more intelligence. Through our solution, that same e-mail has seen transacting successfully elsewhere, linked to known devices, established payment behavior and consistent with digital and behavioral patterns across other customers.

Now we're looking at the same flow, but with a bad actor. Here fraudsters reusing stolen credentials at scale. On the surface, these look like separate transactions, different e-mails, different devices, all appearing unrelated. But our network and sophisticated linking resolves these events into a single identity, not by depending on the device, but by linking across e-mails, locations and behavioral patterns across our network. Even as the fraudster changes devices or spoofs credentials, we're still able to recognize that this is the same identity. These devices and e-mails are no longer isolated events but a singular view into a customer's digital journey.

So now let's apply this in a case study. In this instance, one of our banking customers utilizing our layered fraud solutions, notice a bad actor trying to use stolen credentials to access an account. This same device made multiple attempts to log into the account using different stolen credentials in a short period. In real time, we connected the device, the behavior and network intelligence and flag the behavior as suspicious and inconsistent with a genuine user. And we immediately stopped the fraud attempts and our customer avoided any associated losses.

The scale and visibility of our network enabled us to link that one incident to multiple connected devices and prior fraud activity instantly. From a single suspicious device, we identified 26 additional high-risk devices and blocked 18 more fraud events across 18 different organizations. The result, we were able to stop a coordinated fraud ring that was moving across institutions and channels and additional customers were able to quickly prevent losses. No single institution can see the full picture on its own. However, one incident prevented fraud across our entire network. This example demonstrates how the scale and global reach of our network delivers significant measurable value to our customers.

So I will now turn it over to Matt and Dan to walk through how we create a safer new account opening journey.

Thank you, Kim. My name is Matt Adams, and I'm the Chief Technology Officer and Co-Founder, IDVerse.

And I'm Daniel Aiello, Chief Product Officer and Co-Founder, IDVerse. We founded IDVerse in 2016 and have been with Risk since the acquisition last year. Together, Matt and I lead product and technology for the IDVerse product suite, including the platforms and identity verification capabilities our customers use globally. New account opening has always been one of the most demanding trust decisions in financial services.

Institutions make a binding decision about a new customer with very limited history at the moment of decision. Bad actors only need to succeed once. There is constant commercial pressure to improve quickly because digital growth depends on it. And when something is flagged, the fallback is manual review which is expensive, slow and inaccurate and creates significant consumer friction.

Every safeguard introduced to mitigate risk from CAPTCHA to SMS onetime passwords, document data checks, Q&A, phone calls, add friction to the customer experience and are often insufficient. Financial institutions have long balance 3 competing imperatives: growth, friction and protection. That balance has been fundamentally disrupted by AI. Our own risk intelligence network sees this in the data.

In 2025, synthetic identity fraud attacks tripled within the 12-month period. Fraudsters are producing complete synthetic identities, fabricated documents, fake faces and deep fake videos, breached personal data supplies abundant raw material. Credentials trade on underground marketplaces for as little as $10. Neural network generated fake IDs for around $15. Generic AI tools and frontier models are not designed to detect these threats.

What you're about to see is a recorded demonstration showing how a fraudster or an agent generates a synthetic identity document using a generative AI model. These are fraudulent models hosted on underground sites, sites like this are real and persistent. As demonstration begins, the fraudster selects a country and state and in some cases, with a physical or digital mobile driver license. Genuine stolen or leaked data can be purchased and injected directly, a synthetic face is generated or a real one substituted.

The output in seconds at almost no cost, a very convincing identity document image, sufficient to open a new account at companies without the right safeguards.

The attack method has also evolved. We are now seeing AI agents deployed, prompt injected to act as adversarial networks, attacking the bank's defenses autonomously and at scale. We can now see the fraudster prompt an LLM to target multiple banks and open accounts. Running the IDV process with the synthetic data generated earlier, it is a guardrail demonstration that simulates a real coordinated attack.

As you can see, the agentic AI replicates the human interaction, submitting the ID image to defeat template based checks and presenting a face image or video to spoof liveness. The volume and sophistication of attacks are clearly increasing. To combat this risk, we offer IDVerse integrated with the broader Risk Solutions fraud defense platform.

Here, we're seeing a customer visiting the website of a bank to apply for a credit card. They choose their preferred card and proceed with their application. The first step the bank requires is identity verification. By using IDVerse embedded within its website, the bank can verify the applicants identity while also capturing trusted data to pre-fill the application, reducing friction for the customer.

The applicant taps start verification, which seamlessly opens the IDVerse identity verification flow. They view and accept the privacy consent. They have shown instructions and move to capturing their identity document, we are also able to support a growing list of digital IDs. They confirm the extracted details and present their face for the biometric checks. And within seconds, the user is verified and returned to the bank site to complete their credit card application.

What the customer experiences as simple is anything but.

Behind that short process, multiple layers of proprietary technology operate simultaneously, orchestrated by the neural network, combining physical identity data from the document, biometric intelligence from the face and digital identity intelligence from the broader LexisNexis network. Across these layers sits IDVerse purpose-built neural network, engineered specifically for identity and fraud. This is not a general purpose LLM or third-party AI model. It has been trained for over 7 years on real world fraud attempts, not available in public data sets and is updated continuously as new threats develop.

Let me explain the 3 primary layers of our technology. Layer 1 is document authentication. The physical or mobile digital ID is analyzed using our purpose-built AI designed to detect subtle fraud patterns, document inconsistencies and a wide range of attack vectors and methods seen across our network. It performs up to 300 automated checks, some of these including pixel level analysis, color consistency, lighting angles, micro security features, font integrity and screen detection, along with file-level metadata analysis.

It recognizes virtually all government-issued IDs across more than 200 countries and territories and more than 140 languages. The outcomes are real, trustworthy and present any document with a real identity on it.

Layer 2 is biometric liveness and face match. The face presented by the applicant is analyzed using purpose-built proprietary liveness and presentation attack detection. The system detects synthetic injection attacks, including deep fakes, 2-dimensional and 3-dimensional masks, screen replays and AI generated face swaps, all server-side without requiring additional steps from the user.

Once confirmed live, the face is matched against the document using our own face matching engine, engineered for real world variation and different document standards. The outcome of Layer 2, a live present person, confirmed and matched to their document.

Layer beneath that document and biometric check is Layer 3, which is a context layer, assessing the device, the network, the behavior and how they compare against everything the network has previously seen. Risk Intelligence network was covered in the previous case study. What matters here is what brings the decision. Every applicant is assessed against signals drawn from more than 300 million daily transactions, contributed by institutions across the network.

So every customer benefits from what every other customer has seen. Document, biometric and digital identity, 3 layers each with deep capabilities beneath them. And together, they give our customers the confidence to open good accounts safely.

The impact is measurable. Modern AI-enabled fraud, including deepfakes, synthetic identities and coordinated attacks, are stopped before accounts are opened. Legitimate customers complete onboarding in seconds, manual review volumes fall and the bank can scale digital growth safely. Demand for these capabilities have accelerated across our global customer base since the IDVerse acquisition. As AI enabled attacks scale, institutions are moving to layered defense because their existing tools cannot keep pace. That demand reflects a clear market reality. AI-enabled fraud cannot be met with static general-purpose tools. It requires specialist capability, deep data, expert human judgment and scale that compounds. That is what LexisNexis Risk Solutions provides and what this market is increasingly reaching for.

Now let me hand it back to Rick.

Thank you, Matt and Dan. In summary, we have leading positions in attractive growth sectors. AI is accelerating the volume and the complexity of fraud, which is increasing customer demand for our solutions. We are well positioned to help our customers address these challenges by providing them a better, more holistic view of risk through every interaction they have with their customers, delivering a measurable value uplift.

Our objective is to continue to deliver strong underlying revenue growth in the high single digits for a long time to come, a decade or more, driven by organic product innovation, supported by targeted acquisitions. We are well positioned to continue to adopt new technology to add greater value to our customers, accelerate the pace of innovation and operate more efficiently with underlying profit growth exceeding underlying revenue growth.

We'll now be happy to take your questions.

[Operator Instructions]

The first question today comes from Nick Dempsey with Barclays.

So I have 3 questions. The first one, have your customers so far asked you to work together with some of the big AI modeling companies so that you combine your data with other big processes using AI that are running through the institutions that are your customers? And how do you respond to those requests if you have them?

Second question, can Agentic AI be trained specifically to beat your network and effectively stay ahead of you in terms of fraudulent activity, find the way through all of the sophistication you've been presenting to us?

Third question, there are 12,000 technologists across RELX. I know that's across the whole business, but I guess that's pretty weighted to risk. Do AI tools present an opportunity to make some headcount savings here over time?

Yes, just all through these now. Yes. So have our customers been working with us? I think the first question was, have we been working with our customers with the likes of some of the big AI companies? We're working with our customers to help identify how they want to deploy, how they want to begin accessing our systems, as our risk signals and intelligence into their AI models. We're not there yet in terms of customers actually integrating yet into our systems.

But certainly, the discussions are being had around how do we get access to those signals to inform what our financial services customers are doing to help them better stop fraud on their side. And from the financial crime perspective, the alert remediation space, certainly, they're interacting with us already pulling our signals into their agentic processes for false positive remediation, Level 1 and Level 2.

And then second question, can a Gen AI be trained to find a way to break through Vijay, can you help me?

Yes, certainly. The way I would answer that question is when we talk about the solutions we provide our customers, there's a very delicate balance between a term that we use called 2 terms, precision and recall. Our customers use the same term, but the concept is the same. Precision is a measure of whether we're giving accurate answers without giving spurious answers, right?

So when Rick talked about false positives, that's what we mean. So Agentic AI sitting and talking about LLM does not do that very well. It can be used to augment what we do or we ourselves use to augment what we do. But using Agentic AI in and of itself might cause a problem where, in fact, customers will try to build solutions themselves without our data or without our trustworthy AI they might have a precision problem where they generate lots of false positives.

So for example, in the financial crime and compliance space, that poses a cost problem or an expense problem to our customers. So what do they do? They try to cast a smaller net, whether it's using Agentic AI or using some other LLM and they try to improve the precision. But that causes the opposite problem. It causes a recall problem. So the short answer to your question is Agentic AI in and of itself is not going to compromise the quality of our solutions. You need the data, the breadth of the data assets that we have, along with the domain expertise that we have, along with AI tools we have, all that put together is what renders the value that we offer to our customers.

And I take the third part as well about the 12,000 technologists. Yes. So we are absolutely seeing value in generative -- in AI-assisted coding tools, right? So we are actively experimenting with these. Last year, we saw value, but the value that was generated by these tools is also compromised to some extent because of the technical debt that was creating, meaning it wasn't adhering to our standards. Now because of the evolution of concepts like spec-driven development, we are seeing improved value where not only is AI-assisted coding helping us, but it's also using our tools and our technology stack.

So there is promise. But I will say that while we expect to see some margin improvement over time as a function of better utilization and productivity of technologies, I do think that some of the productivity will be used to bolster our products, improve the quality and security of our products. So it's a mix and match, improve our productivity with the gains and also see some margin improvement.

The next question comes from Henry Hayden with Rothschild & Co. Redburn.

We had 3 on our end. So the first one is on international expansion. You mentioned that you have 45% of your revenues tied to globally applicable solutions. But so far, if you look at the divisional level over the past, let's say, 5 years, there's been fairly limited mix shift in terms of geographic exposure. So we're curious as to how that 45% has evolved over time and how you're thinking about the algorithm going forward from here?

Second question we had was around the moats around the data that piece kind of so your solutions. Fairly comfortable with the moat around data in fraud and ID, but more curious as to that in financial crime and compliance as well as business and credit risk. What level of propriety kind of surround that data? And what prevents a competitor from potentially aggregating it?

And then the third question I had was on customer captivity. So given you're primarily indexed to financial institutions, I appreciate there's quite a degree of captivity around answers need to be accurate. Does that same sense of, let's say, competitive advantage read across to other customer segments as you look to expand there?

Okay. Great. I'll take those. So international expansion. So a couple of points there. Our revenue mix right now is 70%-30% and our global products, those that are unbounded by local data assets is 45%, roughly half and half. The mix has improved on the revenue side, and that's where our international businesses are growing slightly faster than our kind of than our divisional average and slightly faster than the U.S. The U.S. is quite strong as well.

So we're seeing both of those markets grow, and that's why that expansion from when we last spoke, has improved, but maybe not as dramatically as you may have thought. But yes, it continues to see strong movement between taking our global products around the world and getting expansion there. But again, the U.S. is quite a strong market for us as well. So we see strong growth there.

In terms of moat, I think the question was around what is the moat -- you understood the moat relative to our fraud and identity solutions. But let's back up a minute. The data that we use in fraud and identity is the same data that we use across our financial crime and compliance suite as well as our credit risk. So that highly proprietary nature of our public records, our license content, our network data and the analytics and risk insights that we build off of that all goes into our data repository. And that data is used in financial crimes for know your customer and account onboarding.

So a lot of those same insights and differentiation and distinction is applicable to financial crime as well as credit risk. The credit risk data assets, it's all about ability and willingness to pay. And we use the same data asset and build those insights to drive it into the credit risk space. So the moat is equally across all 3 of those sectors. And then finally, I think the last question was around is the -- what was the question?

Relative customer activity from financial institutions versus other customer segments.

Yes. I mean our solutions are the same across customer segments. So the reliability, the accuracy that we build into financial services are also built into those other sectors as well, if that's where that question was going.

The next question comes from Joe Barnet-Lamb with UBS.

You referenced that you have 25-plus contributory and proprietary databases. I think that was in reference to risk overarchingly rather than business services. Is that correct? And if so, how many do you have in business services? And I'm sure this remains a very small proportion of your data sets, but could conceivably drive a significant proportion of the value you create. It sounds like it's a combination of many data sources that multiply the value creation. Is there any way you can frame the influence on outcomes that your proprietary databases have? What proportion of outcomes are touched by proprietary data in some form?

Yes. So the -- you're right. The overall risk contributory database is 25 or so. Business services has 10, and it's an area that we've been really focused on in the past 10, 15 years, driving more and more contributor resources, whether it's network activity or whether it's outcome data supplied back by a customer. So it is a significant source of our data. And in fact, on a daily basis, we're getting more data signals from that data than sort of our licensing and direct sourcing public records data.

So it is a considerable value add. In terms of specific proportions, I can't kind of -- I don't have that number off hand. It's not something we track. But it is a significant value contributor and significantly differentiated. Was there another part to that question?

Well, not really. I think you've given me what you can give me. I mean you did reference that you're getting more data signals than from your licensing and direct sourcing. Could you explain a little bit more what that actually means?

Yes. When you think about what's happening every day with our network activity, we're seeing over 400 million signals a day coming through, and there are multiple elements to that signal. So that contributes to our data repository each and every day, and that continues to grow and grow. And actually, our -- one of our -- in the past 3 weeks, we had some peak days over 500 million. So the signals that we get from those contributory -- and that's just one of these contributory sources is significant and continuing to differentiate and add value across the portfolio.

Next question comes from Christophe Cherblanc with Bernstein.

I had one question about the customer value proposition. Revenues have been growing high single digits. You were mentioning digital interaction going up, I think 13% per annum, attacks going up 15%. So I think that's giving us a sense of the improvement in the value proposition. Even the intensity of attacks is going up, as you were stressing, do you see room to extract a bit more value from what you bring to your customers? And should we expect acceleration reflecting that increased risk exposure for your clients?

Yes. I mean we have our portfolio is roughly $2 billion annually, and we expect to see upper single-digit revenue growth rates with profit exceeding the growth rate in terms of sort of acceleration of volumes there. Certainly, we are seeing more volumes across the digital portfolio with fraud attacks escalating with AI-driven deep fakes and things like that. That all gets blended into the overall mix. So our -- sort of our guidance on revenue growth remains in that upper single digits overall as a portfolio.

But would you say it's fair to assume that the customer value proposition is accelerating versus what you had 2, 3 years ago?

Yes, absolutely. We continue to add more and more capability to the portfolio, continue to expand the value that in the case of ThreatMetrix and our digital solutions bring as well as all of our other solutions bring to the marketplace. So we continue to see strong growth across the portfolio.

Okay. And just one last one related to that. You mentioned that you were a low share of your cost base. How low? Is it way below 1% because based on the client base and the numbers you were mentioning, it seems to be pretty low numbers on an absolute value.

I apologize. I missed the first part of that last question.

Well, you mentioned that your products were a low share of the cost base of your clients. So I was just trying to get a sense of how low the share was? Is it below 1% of the client cost base? Is it 0.1%, 0.5% -- just an order of magnitude would be helpful.

We don't have -- I don't have a specific on that, but it is a very -- it's a low percentage, single-digit percentage if I was to make an estimate.

The next question comes from Steve Liechti with Deutsche Numis.

I've got 3 as well. Just going back to one of the previous questions actually, the sort of consistent growth at 7% to 9%, which is a great growth rate. But given as you kind of alluded to in your previous Q&A, the market and attack growth are growing higher than that, you're innovating very strongly. So I'm just trying to figure out why 7% to 9% is the right number going forward from here? That's the first question.

Second question, I don't think you gave a customer retention number or percentage. Can you give us anything that you can on that for Business Services or broader?

And then the third question is on competition. Apologies, but can you just educate me in terms of who you see your key competitors as being and whether there's been any kind of new innovators in the market that you've lost any share, if you have done to that you might highlight?

Yes, sure. So yes, so back to the revenue growth, we expect to see upper single digits revenue growth for the foreseeable future. Our portfolio is complicated. It's a $2 billion portfolio. So we do have sectors that are growing greater than the average, but offset by some sectors -- some solution sets within the portfolio that are lower growth. So on average, it balances out to that upper single-digit range, and that's where we feel comfortable with.

In terms of customer retention, no, we don't -- we didn't share that, but it is low I guess, the inverse of that, what is our attrition level? It is low single-digit range. Our customers stay with us for a long time. As you saw in one of the slides, where we show the growth with the customer over time. We continue to see -- we land an account maybe with one solution. It could be in FCC screening and then that quickly moves over to fraud and identity and other solutions.

So we continue to grow our customers over time. But we do serve the largest financial institutions in the world, the medium and small-sized accounts as well, where they may be purchasing fewer solutions. And we tend to see -- if we're seeing attrition, it tends to be in those very small accounts.

Second part of the question was -- or the third part. The competition for us is it's generally by geography and use case. So in the U.S., physical identity, there's a quite a few players in the physical identity space and then globally and digital as well -- digital, much less so by FCC, credit risk, front end. So it all depends by geo and by use case. Really, the list of competitors is long, and it's rather -- I rather not get into that level of detail.

But is there anyone who is a major part of the market that's at the scale that you are? Apologies, again, this is my ignorance. Or is the competition more fragmented?

Yes. When I look at the portfolio from what we do from F&I to financial crime to credit risk and business risk that bring the types of solutions that we bring to market, there are very few. There are very few that have that holistic suite that we have. You'll see some on the -- certainly on the credit risk, the credit bureaus are competitors, but we don't tend to play in their traditional credit space. We're doing alternative credit, basically providing underwriting solutions for those that are not on the credit files. So -- and they're partners of ours as well. So it's complementary, a little bit of competitive on the credit risk side.

On fraud and identity, from a digital perspective, physical perspective, certainly, no one has the depth and breadth of assets that we do. And there are players. I mean, bureaus have some components, mostly on the physical side. On the digital side, less so. And then around the world, it just really differs by country. And certainly, there are a number of start-ups that are out there. We see them -- a lot of them are positioning is AI-driven AI delivery engines and things like that. What they don't have is they don't have the depth of the insights that we have from all the risk signals that our solutions provide. So we tend to see them competing on the fringe and not in the main.

[Operator Instructions]

The next question comes from Ciaran Donnelly with Citi.

A couple of questions remaining from myself. Firstly, on M&A, it's been an active piece of the strategy historically. I'd be interested to get your thoughts, do you think the need for M&A to add capabilities given the current pace of technological innovation relative to history has increased? And you can point to any areas specifically that might be an area of interest?

And secondly, one of your peers talked about developing their own proprietary model that in some cases, outperforming the frontier models. I'd be interested to hear is building your own domain-specific model something that you guys have considered? And maybe could you help us think about the cost benefit analysis of model usage more broadly?

So if I heard the M&A question correctly, let me take it. So our whole approach to M&A actually starts with organic product development. First, we're looking to -- we work with our customers to understand what their problems are, what the issues are and how best can we solve those. We then look internally and say, are we -- do we have the capability and the time frames and whether that all works together. And then we say, okay, if that's not -- if that doesn't work for us, let's see if there's a partner solution that we want to consider kind of working with and trying to embed that into our solutions. And that often gives us insight around kind of what a capability gap that we may be missing is all about. And that may be a partner that we eventually acquire or may be a partner that just gave us insight into the capability.

And therefore, we go into the market to then say, okay, we have a capability that we want to solve. And then at that point, we then look for the best in the industry at solving that problem and work with them depending upon timing and things like that to acquire that business. And we've been, as you pointed out, quite successful. It's a continuous part of our strategy.

So as capability gaps do emerge, we then look for that channel approach, then partnership or organic approach first, channel approach and then M&A. In terms of where we're looking, I mean, it's really about use case, FCC, F&I, credit risk. We look across our portfolio, but specific gaps in our portfolio, that's not something that we're prepared to share.

I can take the second part of the question, Rick, about proprietary LLMs. I think that was the nature of the question from Citi. Yes. So it depends on the use case. So when it comes to our people assets or device assets and so on, that is a very deterministic kind of solution. We use LLMs, which tend to be third-party LLMs to augment the quality and scale of our solutions, but it's not really the forefront. There's no need for us to build a proprietary LLM. In fact, it could be our solution.

But when it comes to IDVerse, that is, in fact, a proprietary model. It is a proprietary element that we built, a neural network we built, I should say. So when you heard Matt and Dan talk about the IDVerse solution, that is a neural network that we built ourselves starting in 2016, trained with data that has been so over the last 10 years, it's been trained because it's a very specific kind of use case that operates on liveness detection, document authentication. And then we take the output of that and feed that into our risk intelligence network. So it really depends on the use case. We absolutely do build our own domain-specific LLM neural networks as we see fit.

This concludes our question-and-answer session. I would like to turn the conference back over to Rick Trainor, CEO, for any closing remarks.

Yes. Thanks. Yes, I'd like to thank you on behalf of the team for taking the time to join us today. I hope you share our enthusiasm for the business with its leading positions in attractive grreasing customer demand for our solutions, which gives us confidence in our objective of continuing to deliver strong underlying revenue growth in the high single digits for the foreseeable future. Thanks, and have a great day.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

Good morning, everybody. Thank you for taking the time to join us today. As you may have seen from our press release this morning, we delivered strong financial results in 2025. We made further operational and strategic progress, and we continue to see positive momentum across the group.

Underlying revenue growth was 7%. Underlying adjusted operating profit growth was 9%, and adjusted earnings per share growth was 10% at constant currency. All four business areas continue to perform well. On this chart, you can see the relative sizes of the business areas and their growth rates with underlying adjusted operating profit growth exceeding underlying revenue growth in each business area.

In Risk, underlying revenue growth was 8% and underlying adjusted operating profit growth was 10%. Strong growth continues to be driven across segments by the development and rollout of our deeply embedded AI-enabled analytics and decision tools with over 90% of divisional revenue coming from machine-to-machine interactions.

In Business Services, which represents over 40% of divisional revenue, strong growth continues to be driven by financial crime compliance and digital fraud and identity solutions and strong new sales. We continue to expand our differentiated data set, build out our global fraud infrastructure and more deeply integrate advanced authentication and behavioral intelligence.

In Insurance, which represents around 40% of divisional revenue, strong growth continues to be driven by innovation and adoption of contributory databases and market-specific solutions, supported by positive market factors and strong new sales.

We continue to expand our products across the insurance continuum and across the insurance lines, while adding data sources and analytics to enhance value for our customers. Going forward, we expect continued strong underlying revenue growth with underlying adjusted operating profit growth exceeding underlying revenue growth.

In STM, underlying revenue growth was 5%, and underlying adjusted operating profit growth was 7%. Improving momentum is being driven by the evolution of the business mix towards higher growth, higher value analytics and tools supported by the increasing pace of new product introductions and strong new sales.

Databases, Tools & Electronic Reference, which represents around 40% of divisional revenue, delivered strong growth, driven by higher value-add analytics and decision tools and we continue to expand our solution set built on our industry-leading trusted content with an ongoing series of new releases.

In Primary Research, which represents a little over half of divisional revenue, good growth continues to be driven by volume growth. The number of articles submitted continued to grow very strongly across the portfolio by over 20% in 2025, and the number of articles published grew 10%. Going forward, we expect good to strong underlying revenue growth, with underlying adjusted operating profit growth exceeding underlying revenue growth.

In Legal, underlying revenue growth improved to 9%, with underlying adjusted operating profit growth of 12%. Strong growth continues to be driven by the ongoing shift in business mix towards higher growth, higher value legal analytics and tools.

In Law Firms & Corporate Legal, which represents around 70% of divisional revenue, double-digit growth is being driven by continued adoption of our core AI-enabled legal platform and integrated Agentic assistant, Lexis+ AI and Protege. Ongoing releases of new functionality and deeper integration with our comprehensive, verified legal content is enabling us to increase our value add and serve an increasing number of use cases.

Going forward, we expect continued strong underlying revenue growth, with underlying adjusted operating profit growth exceeding underlying revenue growth.

Exhibitions delivered strong underlying revenue growth of 8%, reflecting the improved growth profile of our event portfolio and good progress on our growing range of value-enhancing digital initiatives. Underlying adjusted operating profit growth of 9% was ahead of revenue growth with margins now significantly above historical levels. Going forward, we expect continued strong underlying revenue growth with an improvement in adjusted operating margin over the prior full year.

Our strategic direction is unchanged. Our improving long-term growth trajectory continues to be driven by the ongoing shift in business mix towards higher growth analytics and decision tools. This is being supported by the continued evolution of artificial intelligence, which is enabling us to add more value to our customers as we embed additional functionality in our product and to develop and launch products at a faster pace.

Our revenue growth objectives for the business areas remain: For Risk, to sustain strong long-term growth; for both STM and Legal, to continue on their improving growth trajectories; and for Exhibitions, to sustain strong long-term growth. When combined with continuous process innovation to manage cost growth below revenue growth, the result is a higher growth profile with strong earnings growth and improving returns.

I will now hand over to Nick Luff, our CFO, who will talk you through our results in more detail. I'll be back afterwards for a quick wrap-up and Q&A.

Thank you, Erik. Good morning, everyone. Let me start by providing more detail on the group financials. As Erik said, underlying revenue growth was 7%, with underlying adjusted operating profit growth ahead of that at 9%. As a result, the adjusted operating margin improved by just under 1 percentage point to 34.8%. The strong operating result flowed through to adjusted earnings per share, which at constant currency increased by 10%.

Cash conversion was again strong at 99%. After acquisition spend of GBP 270 million and the completion of the GBP 1.5 billion buyback, leverage ended the year at 2.0x at the lower end of our typical range. Given the strong overall performance, we are proposing an increase in the full year dividend of 7% to 67.5p per share.

Looking at revenue, you can see how all 4 business areas contributed to the overall 7% underlying growth. As we discussed at the half year results, we have separated out the reporting of print and print-related revenues and profits, reflecting changes to how we manage the distribution of print versions of our content. The proactive steps to reduce our involvement in print-related activities continued in 2025, resulting in a reduction in associated revenue of over 20%.

For the group as a whole, total revenue growth at constant currency was 4% after the portfolio effects in Risk, Legal and Exhibitions and after the step-down in print activities. In addition, there were cycling effects in Exhibitions with 2025 being a cycling out year. In sterling, total revenue growth was 2% impacted by the relative strength of the pound against the dollar compared to the prior year. Here, you can see the 9% underlying growth in group adjusted operating profit.

As Erik mentioned, we continue to manage cost growth to be below revenue growth in each business area. As a result, Risk, STM and Legal each delivered underlying profit growth 2 or 3 percentage points ahead of underlying revenue growth, while Exhibitions was 1 point ahead, reflecting a better cycling in the year.

The profit contribution from print and print-related activities declined but at a lower rate than revenue. As I said at the half year results, going forward, we expect profit from print and print-related activities to continue to decline in the high single digits each year in line with historical trends.

Portfolio effects and the decline in print were a slight drag, leaving total adjusted operating profit growth in constant currency at 7%. There was a similar currency effect on profit as there was on revenue, giving adjusted operating profit growth in sterling of 4%.

With profit growth ahead of revenue growth, margins improved across all 4 business areas, driving the overall improvement of 90 basis points to 34.8%. Margins were up by 40 basis points in Risk, 70 in STM and 80 in Legal. Exhibitions margin increased by 250 basis points, aided by prior year disposals and the effects of cycling.

Turning to the group adjusted income statement. You can see here the underlying growth was 7% in revenue and 9% in operating profit. The interest expense was slightly lower, with the decrease reflecting lower average interest rates partly offset by higher average debt balances.

The effective tax rate was 22.5%, in line with the prior year. Net profit was up 8% at constant currency and up 5% in sterling to over GBP 2.3 billion.

With the lower share count as a result of the buyback program, adjusted earnings per share were up 10% at constant currency and up 7% in sterling to 128.5p.

Turning to cash flow. Cash conversion was strong at 99%. EBITDA was over GBP 3.8 billion and CapEx was GBP 525 million, equating to 5% of revenue. After interest and tax, total free cash flow was over GBP 2.3 billion.

And here's how we deployed that free cash flow. We completed 5 small acquisitions with total consideration of GBP 270 million and made 2 small disposals. The most significant acquisition was IDVerse, an ID document verification platform for business services in Risk, which completed in the first quarter of the year. Dividend payments were GBP 1.2 billion, and as I mentioned earlier, we completed GBP 1.5 billion of share buybacks.

Overall, year-end net debt was GBP 7.2 billion. Including pensions, the ratio of net debt to EBITDA calculated in U.S. dollars was 2.0x at the lower end of our typical range of 2 to 2.5x.

Our priorities for the use of cash remain unchanged. Organic development is our #1 priority with CapEx consistently around 5% of revenues. We augment that organic development with selective acquisitions with this level of spend typically being the most significant variable in our uses of cash, depending on the opportunities that arise. Average acquisition spend over the last 10 years has been around GBP 400 million per annum with 2025 a little below that average.

We pay out around half of our adjusted earnings in dividends and have increased the dividend every year for well over a decade. Leverage has typically been in the 2 to 2.5x range. Strong cash generation, improving EBITDA and modest acquisition spend in the year mean that leverage at the end of 2025 was at the lower end of that range.

We continue to return our surplus capital through the share buyback with GBP 2.25 billion of spend announced today for 2026, of which GBP 250 million has already been deployed.

With that, I will hand you back to Erik.

Thank you, Nick. Just to summarize what we have covered this morning. In 2025, we delivered strong financial results, and we made further operational and strategic progress. Going forward, we continue to see positive momentum across the group, and we expect another year of strong underlying growth in revenue and adjusted operating profit as well as strong growth in adjusted earnings per share on a constant currency basis.

And with that, I think we're ready to go to questions.

[Operator Instructions] We take the first question from the line of George Webb from Morgan Stanley.

I have got a couple of questions, please. Firstly, big picture one, it's hard to miss the kind of broad concern or fear that's happening across a lot of stocks today. If we pick up on your Legal segment, I guess the latest one of those worries is a concern that you might face incremental competition around AI-enabled workflow tools from other large software companies.

Maybe if we take one step back, for the last couple of years in Legal, we've seen you talk about product launches which use Gen AI, more product adoption by customers and therefore, underlying acceleration in the Legal business.

I guess the question is, do you or have you seen anything in your business in terms of lead indicators or numbers on product adoption, conversations you're having that calls into question your ability to continue to participate in that tech adoption cycle, and that means we should be thinking about potential deceleration in legal before any potential further acceleration? That's the first question.

Secondly, just on STM, given the slight bump in the outlook there. On one hand, you talked to kind of the strong submissions growth and maybe the early ramp of new products such as LeapSpace, but then I guess the full open access growth might moderate in the mix this year, the U.S. funding environment is still a little bit tough. Could you maybe just outline some of those growth considerations in the mix for 2026?

Okay. Well, maybe I'll have -- thank you. Maybe I'll ask Nick to comment on the specifics on growth, adoption, penetration, rollout usage on Legal. And then I'll comment on that a little bit and move on to the second.

George, I mean I think the opposite. I mean, we see these tools as adding value, enabling us to build the functionality into our products. And you're seeing that come through in the adoption, the usage. And if you look specifically at the Legal business and Lexis+ AI, the enterprise-wide subscription customer base has more than doubled in the past year.

And the usage is going up faster than that. We have users in the multiple hundreds of thousands now across the globe on Lexis+ AI. We're seeing strong demand for what we do with the product built on that trusted curated content set, it remains very important to the customers, and these tools are enabling us to add value and grow faster.

I think just if you back up a little bit to your broader question about workflow software, I think it's important to remember that the core of our strategy always starts with our uniquely differentiated, comprehensive content, our collection of trusted, verified, continually updated content and data sets. And we then leverage our deep understanding to combine these content assets with sort of advanced evolving technologies and these evolving AI tools to deliver increased value to our customers.

And I think it's important to understand that we have worked with this strategy inside Risk with the evolution of AI tools, extracting machine learning tools for over 15 years, and that's been the core driver of the whole evolution of the Risk business to now being 40% of our profits, growing 8% a year on revenue, and this year, 10% on profit.

And we have had the same technology-agnostic philosophy and tool-agnostic, multi-model architecture from the beginning of the Gen AI trends for over 3 years. We've been partnering closely with all the large language models providers, including Anthropic and OpenAI for that time period. And as they continue to build out their models and tools, we continually evaluate all the new releases, including often through previews as a partner, and we often test them through ongoing interaction with our customers to determine if they can help us add more value to our customers if we embed them in our tools.

So any new tool that you read about, hear about, we're probably already testing it, involving it in our platform and seeing if we can add more value on our platform to our customer value equation. And often, as you say, there are several companies out there that are developing workflow tools that effectively are today serving -- they're trying to serve or starting to serve some of the use cases that other software companies are serving today.

In Legal, large law firms typically use over 100 of these software companies for different workflow tools, different admin procedures. And if those tools embedded in our core content platform help our customers add more value, we will embed the best of those new tools into our platform and act as an integrator of those and make them work with our customers.

And if they're not relevant to the content-related use case, the content-related workflow and if it's just workflow that's today being served by software companies, then we don't integrate them directly.

We often look at alternative ways to be interoperable and compatible with them so that our content sets, our deeply differentiated content set on our content platform can actually be accessed in the different workflows and we believe that, that way then we enhance the utility of, and therefore, the value of our platform if it can be accessed in workflows where people are more efficient and more productive and in the area where we don't want to be or operate ourselves.

I mean today and historically, we have virtually no revenue in any of our divisions from what I would describe as workflow software-related services.

And sorry, to the STM question. I mean you asked about submissions and publication volumes, George. The fact is that science remains a totally global industry. The number of scientific researchers in the world continues to go up. The information intensity of science continues to increase. The desire and the speed at which people want to be published continues to increase.

And so we -- as you saw in the -- we had strong growth in submissions last year, over 20%, the number of articles we published over 10%. And that has not slowed down. We're seeing that continue into this year. There's continued strong momentum in primary research. And there's always in any one country, there can always be things happening. But if you look at it in an overall sense, we continue to see strong demand for primary research publishing.

We take the next question from the line of Nick Dempsey from Barclays.

I've got 3. So first of all, for the Protege AI workflows, which you are now starting to roll out, can you please talk through what differentiates those offerings from the competition in that broad AI workflow market in a bit more detail, please?

Second question, there have been some concerns knocking around about autonomous driving and the auto insurance market. Can you talk about your exposure, the impact as the auto market shifts gradually towards autonomous driving and give us a sense of whether you see any long-term risks around that?

And number three, when you refer to strong new sales in 2025 for the group and then in Legal, you'd say renewals and new sales are strong across all 3 segments, am I right in thinking that those new sales will have only a very modest effect on '26 growth, but you're signaling that they should be supporting growth through '27, '28 and beyond?

Yes. So I'll let Nick to start with the first one.

Yes. So I mean, the big difference between what Erik was touching on earlier, all the things we're offering to do is the content that's behind them. We would describe what the workflow tools that we're introducing as being content-enabled, and that's a key differentiator.

It's not that other tools can't be useful to people. And as Erik touched on, many tools are used by lawyers and other professionals. But the ones we have, if you're actually doing anything that relies on trusted curated content, then that's where the differentiation comes in.

We also, of course, have the advantage of the customer understanding and the sheer scale at which we already operate. As I touched on earlier, we have hundreds of thousands of users of Lexis+ AI. And so we can see how it's used, and we can see what's useful and constantly be updating the quality of the answers that we're able to provide, and that's a key differentiator as well.

Yes. Yes, I mean I just want to add to that, but I think it's important to look at this is that the workflows we're developing, I think when we first released Protege, we were talking about order of magnitude sort of 50 workflows or so in earlier, and these have been released out in phases, continue to be released out in phases and upgraded as we go along.

At the moment, we're probably nearing 300 different workflows -- specific workflow tools. And we can develop these on our content, on our platform and launch them to our customers at the rate of probably another 2 or 3 a day in this machinery that we have.

But again, these are content-related workflows that are embedded in our platforms that help add value to our customers the way they operate with us and it's unrelated to the kind of industry that is the broader legal tech software industry where people are spending money on software or workflow solutions for operating an admin. And that's where we separate the two and try to be embedded with the first category and be interoperable with the second category.

As you know, we're fully embedded in Microsoft since many years ago for our customers, they can fully operate and work between our tools and the Microsoft tools. That does not mean we're trying to compete with them or operate Microsoft general admin workflows in any way.

But it enhances the value of our content and our utility of our platform when our content-specific workflows fit right on our content, but it also enhances the value when you can use our LexisNexis AI-related platform and workflow interoperably with Microsoft, for example. And we have about 25 of these different existing partnerships in Legal today, and I'm sure there'll be many more in the future, yes.

So Nick, on the autonomous driving question, obviously, there are lots of trends affecting the auto insurance industry all the time. Enhanced safety features is part of that, automatic braking, telematics, some autonomous driving.

And I think we see that as the whole industry evolving to make driving safer, generate more data and everything becoming more complex as you do that. And in that environment, what we do where you get sophisticated risk analysis, combining the data from -- about the driver, about the vehicle, about how it's been driven, the interaction between cars being driven in different ways, that just creates opportunity for us.

The value at stake actually goes up, and it's been a trend for many years that you get fewer accidents, but the severity of them and the cost of them goes up. So the value at stake actually is getting higher. And in that environment, I think we're extremely well placed to add more value because of the additional data and analytics that we can provide.

And your last question, Nick, was on strong new sales. You're absolutely right. I mean obviously strong new sales. New sales are -- in a subscription -- heavily subscription based business as we are, they are only a small component of what's relevant to the current year revenues, but they are a good indication of the momentum there is in the business and ultimately, what drives the long-term growth trajectory, and that's why we're flagging this morning.

The next question comes from the line of Christophe Cherblanc from Bernstein.

I have 2 questions. The first one is on STM. I guess, we have a sense of the lawyer population, but it's harder to understand the addressable population for tools like LeapSpace. So I was curious whether you had any number in mind or any number of institutions and how long it would take to ramp up penetration?

And the second question was about pricing. I think you've been insisting that especially in Legal, you are no longer pricing per seat, but I was curious as to what was the extent to which you've been changing pricing contract over the last 12, 24 months and whether you intend to further adjust pricing going forward?

Yes. So on the STM side, we are launching several different tools into that market, as we've told you. Several tools have been going for now up to -- well, 1 year or up to 2 years in some instances, and we continue to see what the value uplift is to the customer, what the usage growth is and what the user growth is and usage growth, and we can see the value they're getting.

From the new forward-looking LeapSpace launch, which has just recently launched commercially, we can see that is a significant value uplift to the users, several of them report very significant time savings or productivity gains or improved results from specific use cases that are very material.

And we look, therefore, at the potential addressable market as being basically all the institutions that today have any of our platforms in use, right, or any of the subscribers. And that order of magnitude is in the thousands. I mean, it's over 10,000, depending on when you want to define it, somewhere between 10,000 and 15,000 institutions, right, as potential institutional customers.

When it comes to individual users, which also in the end could be a customer for this, I would look at it as is typical that people refer to the total number of researchers in the world at somewhere a little bit above 10 million. That's the scale of this.

And if you look at the question of how do we price them, our approach here is to price this platform based on scale of institution and research intensity of the institution. So therefore, there's a set of pricing metrics regarding what type of institution it is.

We are also likely to, over time, come up with an individual researcher subscription option for those researchers who operate in a different way that they should -- that might want to access the capability of this in their daily research life. But we're very early stages on the commercial side of this, and it's sold and priced separately from our other content tools.

But the indication we're getting from our customers, the feedback we're getting in terms of the value adds and the excitement is very strong. But as you said, everything in the STM industry goes a little more slowly than it does in other industries, partly because of how they think of funding and spending and budget and also because the purchase cycles, the decision cycles at academic institutions are typically slightly more involved and take longer.

But we are very positive on the ability for this platform to continue to add value to our customers and meaningfully impact our long-term value-add and growth trajectory in this division, but it's going to come through very gradually.

We take the next question from the line of Thymen Rundberg with ING.

Two from my side. I have one on operating leverage and margins. So you've done a great job in managing cost growth below revenue growth in the last few years, also 2025, profit margins are expanding nicely.

As we are now moving in more compute-intensive AI or agentic workflows that just basically require deeper reasoning, how are you leveraging your scale and your -- what you've just talked about as well, your model agnostic approach to ensure that you can still drive that margin expansion while delivering these more sophisticated capabilities?

And then the second question is with the pace of this AI and agentic AI innovation across all your divisions, I was wondering if you could walk us through a bit how you're currently assessing the balance between returning capital via buybacks, what you've now increased, and more or perhaps larger strategic acquisitions.

And so given that your leverage remains at the low end of your 2 to 2.5 range and organic investments are still your priority, I was wondering if you could highlight just when does it make sense to use the balance sheet a bit more actively, particularly in light of competitive dynamics?

Thank you for that. I'm actually going to ask Nick to tell us about both of those.

Yes. I mean obviously, the new technologies that are evolving are giving us great opportunity to build additional functionality in our products, but they're also giving us the opportunity to improve our own processes, make our own processes more efficient.

So we're using those to -- internally, which enables us to get to market faster, but also ensure we can keep cost growth below revenue growth. And I don't think -- obviously, we're spending more on some things than what we're spending with large language model providers, et cetera, as our customers use our products more and as we use those technologies more.

But equally, with other things that we can do more efficiently than we couldn't before. And there's nothing we see in the overall dynamic that means we can't keep cost growth below revenue growth. And if anything, as we touched on in the outlook statements, the gap between profit growth and revenue growth can be -- potentially be a little bit wider. And that's just through cost control and the opportunity that it's -- that these new tools are giving us.

I think -- your second question, I think, was about acquisitions and balance sheet and how we might use it. The primary focus remains on organic development. We have the skills and the opportunity. We have all the assets we need to innovate and bring new products to market and value to customers using that. We will look at acquisitions where we see the opportunity -- where we see something that can enhance and accelerate what we're doing. But they have to fit, they have to fit with what we're doing.

And obviously, with those specific criteria, there's only a few things that are available at any time that makes sense. We could -- and we've had a couple of -- a few years now of relatively low M&A spend. That's not deliberate. It's just the way things have -- what's come up and it's perfectly possible that in the next period, we may see slightly 2 or 3 larger acquisitions come up, and we would absolutely invest in those if we saw the opportunity, but it's not the core of the strategy. The core strategy is organic.

And in terms of where the leverage is, as you rightly point out, because we've had relatively low M&A spend in the last couple of years, we're at the bottom end of our leverage range. Clearly, we reflect that when we think about the buyback, and we have announced a buyback of GBP 2.25 billion this morning, which is up 50% from the buyback in the previous year.

That -- if you take the sort of average M&A spend we have for the last few years, it's been around the sort of GBP 250 million mark, then all things being equal, that would put us roughly in the middle of our leverage range of 2 to 2.5x. So that's why it's been pitched at that level.

We take the next question from the line of Ciaran Donnelly from Citi.

Firstly, just in terms of Legal, can you help us understand the mix between publicly available data and proprietary created curative data that underpins those products? And perhaps just comment on how difficult it will be to replicate those data sets, just looking to get a sense of how deep that competitive moat is?

In addition, can you just clarify with regards to your comments on interoperability, would you be open to licensing use of your proprietary data to be integrated into, I don't know, API plug-in such as Claude Cowork?

And then lastly, just in risk, it looks like the base market growth contribution was a smaller contribution in 2025 versus '24. So can you just help us understand the dynamics there? And looking forward to 2026, what the mix of growth from base and product innovation is likely to be?

Yes. So let me start with the question of our content sets. As you know, we describe RELX as a global provider of information-based analytics and decision tools. And everything we do is built on that information base, which is a foundation of unique and comprehensive content and data sets. And that applies to all our divisions.

And our assets are both historically comprehensive and continuously updated on an industrial scale across our divisions. And in each one of our divisions, it includes some form of public records accumulated over decades, some of which are no longer publicly available, some of which are theoretically public, but extremely difficult and complicated to collect because of the format or in print or in different locations.

Then they also include licensed data sets. Across the company we have licensed data for over 10,000 different sources. Some of those sources, the usage is regulated and controlled, and we can only use them in certain ways in our tools. We then have unique contributory data sets, and we have some of those involved in Legal as well.

And we have dozens of those contributory databases across the company. We then have proprietary data and content that we have created ourselves, written ourselves, either within our pool of internal employees or external contractors have created them for us over many years, right?

But we combine these content and data sets then with our deep customer understanding to build proprietary algorithms, judgment, inferences and interpretations, which accumulated over decades, delivered unique insights and significant value to our customers themselves, and this would be extremely hard, if not impossible, to replicate to the same level of value.

And this is what we mean when we talk about the fact that we have a content advantage that we believe is very sustainable and very strong and are very high value to our customers across our divisions, including Legal. So if you then look at the question, will we consider just licensing out our content sets and on? No, this is a centerpiece of our strategy. This is what we are.

We are an information-based company. We're a content-based company. And everything we do is built around that unique, comprehensive information base. And that's the foundation for our product today. It will be the foundation of our products and their value-add in the future.

And is it possible some [ small sliver ] in some noncore areas could be licensed in some places? Yes. We've always done copyright sales here and there for decades, but that's not material. It's not the core of our strategy.

Our core -- the core of our strategy is to leverage those deeply embedded content and data sets and embed these new tools on top to enhance the value of those content platforms to our customers. And that's what we're seeing a confirmation of when we do that to our customers, we see that they see a value uplift.

We see the spend uplift they are willing to go on those because they see the higher value. We see that customers do that when it rolls out. We see that the users, we have more active users on the new higher value-add platforms and that they use them more.

And I think your last question was about the split of the risk growth, the 8%. As you rightly pointed out, the contribution from new products has gone up. This year, the split was 6% from new products, 2% from older products compared to 5%, 3% the previous couple of years.

I wouldn't read too much into that. It's only a small shift. If anything, it just demonstrates that the pace of innovation has increased. The older products perhaps are being replaced slightly quicker with new products, new functionality. And therefore, the split has shifted a little bit, but I wouldn't read too much into it.

We take the next question from the line of Steve Liechti from Deutsche Numis.

I'll take 3 well, please. First of all, just relatively simplistically, just if I'm a lawyer, and I've now embedded Harvey or Legora into my workflow, just why am I going to buy Protege as well as a workflow tool, maybe put that in the context of a large lawyer and a small lawyer? So that's the first question.

Second question is on STM. You've given your -- you've moved your guidance from good to good to strong like-for-like growth. Is that code for saying that you think like-for-like is going to go from 5% this year to more like 6% next year?

And then the third question is on the Risk. Just we're having a lot of conversations with people on the kind of disruptive stuff going on in the market. Just remind us or rehearse the arguments on why an LLM or disruptor would find it very, very difficult to break into the risk market in terms of either the business services bit or insurance?

Nick, would you like to take the first one?

Yes. Look, there are obviously various tools out. And as we said before, the whole ecosystem in which lawyers operate, they've traditionally used all sorts of different tools for different functionalities. It does depend on what sort of work you're doing.

But if you're doing work that -- legal research work, in particular, but anything that it relies on content and what the latest information is, the latest law is, then as we've been outlining, we have a significant competitive advantage because of the data set that we've put and the content that we've articulated a couple of times already on this call.

That doesn't mean to say that lawyers won't use other things as well. And if they're good tools, then as we said, we'll look to see whether we can use that functionality, replicate it in our products or make it interoperable with our products. And we'll continue to do that.

But I think we're clear that we have a big customer base already using our Lexis+ AI with Protege tool that runs into the hundreds of thousands of users, tens of thousands of customers, so the scale of what we're doing is already way bigger than a lot of things -- lot of other things that are out there. So I think the starting point with that content advantage is very good for us.

And I think it's important to distinguish here between content players and competing in content, which is what we do with these layers on top, which is content-enabled processing that adds value to the content, and the people who are building workflow tools that are not in the content business at the scale that we have or the comprehensiveness of the historical trust and verified content that we have.

But there, there are several hundred software and workflow companies ranging all the way from Microsoft at the top to very specialized tools that are used by lawyers in many ways. And as we said, many of the large law firms have 100 of these different tools. And the two that you mentioned that are coming up that for workflow tools that enable processing and workflows, they are more the way they describe it, going after that much larger software and services market in the legal tech space.

And in a way, they have explained that they see that their biggest threat to their in their quote publicly is the LLM tools and LLM-related workflow tools themselves. We see them as additional partners. We're partnering with already 25 of these workflow and software-related companies in that space, and there's nothing that says that, that couldn't be -- do more over time. So we see them more as complement than competitors.

Steve, your second question was about the guidance around STM. I think as we said in the statement this morning, we have got improving momentum in STM. We are seeing an increased pace of the introduction and rollout of new products. We can see it in the strong new sales. So the business is in very good shape.

Clearly, it's a very heavy subscription business. So things tend to change relatively slowly. But without getting into the numbers, clearly, the outlook statement is a more positive statement than we've had previously. That is an upgrade in our outlook.

And your last question was about the Risk business and LLMs and things. I think the most important thing to remember about the Risk business is 90% of its revenue comes from machine-to-machine interactions.

And this is a massive scale of the data sets we have and the data we collect from all the -- as we've outlined a couple of times already on this -- in this discussion, the thousands of sources, the public records, the contributory data coming back from customers with that network effect that they can all benefit from, what we do with the data, the algorithms that we apply to it and it's incredibly difficult to replicate.

It's a heavily regulated area, what data you can collect, how you can -- how you're allowed to use that data is heavily regulated. And I think given it's almost all machine to machine, I think we see lots of opportunity to continue to use new data sources and using technology. But I think we will be the beneficiaries of that.

But I think it's important also to point out that Risk has been at the forefront of using AI technology now for close to 20 years. The core driver behind the entire growth rate and the growth improvement over the last 15 years in Risk has been the fact that we have all these unique comprehensive data sets that most people don't have access to any of those particularly the contributory data sets and some of the internal data sets that we generate in those markets.

But the real enabler has been the fact that we have had a technology-agnostic philosophy for that entire time period and continuously look at new AI and machine learning tools and new algorithms for a very, very long time. And whenever anything comes out that can help us increase the value to our customers, we have tested them and embedded them. And that's why at this point, we are a 90% embedded machine-to-machine AI-enabled algorithm business.

The new or evolving generative AI tools actually do not add significant value to those kind of mathematical calculations. I mean, just to give you an illustration, in one of our contributory database offerings, we process around 400 million transactions per day in a mathematical continuously improving model, right? So this is a completely different type of business that went through the AI enablement transformation starting about 20 years ago, it started and that's continuing to evolve, and it's already very, very far down this path.

I mean I could remind you that it's exactly 20 years ago this year that because of how we approach big data, data science and algorithms, we picked up knowledge of what was going on over at -- no, over at Palantir, yes, exactly over at Palantir. And I went out to visit them personally about 20 years ago and talked about how our different technologies compare and how well we could do together and some of their tech people were at our conferences and so on.

And we've evolved into a high-volume, algorithm-driven very low price per unit, but very high volume sort of transaction-based pricing installed inside industries, and they've evolved in a complete opposite direction, but we still leverage the same technology heritage and the same thinking and approach to big data, data clients and AI.

So this is not a new thing, and it's not something that's likely to impact the trajectory of the Risk business in any way other than continue on the path we've been on to evaluate and look at and embed any new possible AI tools from any source that can increase the value to our customers of those algorithms we operate today.

[Operator Instructions] We take the next question from the line of Henry Hayden from Rothschild & Company Redburn.

I have 3 from my end. The first one on STM, how do you think about the corporate opportunity? So it's one you discussed in the past as a large addressable market with attractive structural growth profile.

We were hoping for any incremental color you could give around end client preferences. And if there's a different approach that needs to be taken in going after that opportunity in terms of product functionality? And is there an appetite to grow corporate within the mix? And if so, what unlocks better exposure to that underlying growth?

Secondly, within Legal, we're seeing this structural uplift in tech investment from law firms, which adds support to your growth, but also can drive some degree of experimentation for new solutions around legal research and workflows. At what point would you expect firms to kind of consolidate how many products they're taking? And how do you think about your positioning against that consolidation?

And then finally, on Risk, you called out strong new sales and insurance again now. Is there a specific product or line item driving this? And are those competitive displacements? Or is there something else at play here?

Well, I can address first the STM market. The corporate market is, we believe, an important future growth opportunity for us. It is a relatively small segment of our revenue today. And we believe that it is more commercially oriented, and as these tools that we build become higher value, more usable with new tools on top of our content that we see an opportunity to continue to sell and package those in a way that is more appropriate for the corporate market.

We believe that we're going to continue to see that growth rate there pick up as well over time as those tools are developed, integrated to add more value. But it's a relatively small segment today. It's likely to be gradual, even though on some of the tools we've rolled out today, we've actually slightly faster uptake on the sales cycle than we do in the academic markets as early signs. So we're positive, but it's small and it's still going to be gradual. On the Legal tech?

So Henry, on the legal tech. And look, I think law firms will continue to evaluate and look at new technology and look at new tools. The legal research market clearly is very consolidated already with, obviously, the two big players, of which we're one.

But if you look at the wider technology provided to law firms, which is a big market, and all commentators think that's going to grow quite significantly. Individual law firms may choose different strategies, but I'm sure they'll continue to experiment. And we think we have a strong offering to move into some of that market and to continue to add value in that more consolidated legal research market where we play.

And your third question was on insurance and new sales. That business is going well. It is -- we continue to innovate. We continue to have new sources of data, and we touched on it earlier when we're talking about data coming off vehicles, from vehicles, about vehicles. For example, new identity data being brought to bear. We are using new data sources in different lines of insurance.

So for example, using electronic health records for -- in the life insurance market, using aerial imagery or video taken inside the home, analyzed by AI to inform property. And these are additive. These are additive to what's already there. This is not typically displacing anything. It's -- these are not either/or type products.

It's something that functionality and analytics that wasn't available before. And as we innovate and make it available, then it comes into the marketplace and helps the insurance companies become more efficient, helps them price risk more accurately and they see value in them, and that's what's driving the take-up.

As there are no further questions from the participants, I would like to turn the conference back over to Erik Engstrom, CEO, for any closing remarks.

Well, thank you so much for taking the time to join us this morning. I appreciate you listening to us and asking us questions. And I look forward to talking to you again soon.

Good morning, everybody. Thank you for taking the time to join us today. As you may have seen from our press release this morning, we delivered strong underlying revenue growth of 7% in the first 9 months, and we continue to see positive momentum across the group. Our improving long-term growth trajectory with a higher-quality growth profile continues to be driven by the ongoing shift in business mix towards higher growth analytics and decision tools that deliver enhanced value to our customers.

The full year outlook for this year is unchanged, both at the group level and for each of the 4 business areas. In Risk, underlying revenue growth was 8%. In Business Services, which represents over 40% of divisional revenue, strong growth continues to be driven by financial crime compliance and fraud and identity solutions with strong new sales. In Insurance, representing around 40% of divisional revenue, strong growth continues to be driven by the expansion of solution set, positive market factors and strong new sales. In STM, underlying revenue growth was 5%, with developing momentum supported by the increasing pace of new product introductions and renewals and new sales ahead of prior year.

Databases, tools and electronic reference, which represents around 40% of divisional revenue, delivered strong growth. And the recently announced next-generation end-to-end AI-powered restructure solution has received very positive feedback. In Primary Research, which represents a little over half of divisional revenue, good growth continues to be driven by very strong volume growth with article submissions growing by over 20% and articles published growing by 10%.

In Legal, underlying revenue growth was 9%. Renewals and new sales remain strong across all key segments. In law terms and corporate legal, which represents around 70% of divisional revenue, double-digit growth is being driven by the continued success of Lexis+ AI, our integrated generative AI platform.

Protégé, our next-generation AI legal assistance, continues to see rapid growth in usage and further expansion of use cases. Our most recent launch, Protégé General AI, has been very positively received.

In Exhibitions, underlying revenue growth was 8%. The continued strong growth reflects the improved growth profile of our Events portfolio and good progress on value-enhancing digital initiatives.

To summarize, we delivered strong underlying revenue growth in the first 9 months. We continue to see positive momentum across the group and an improving long-term growth trajectory with a higher quality growth profile.

And with that, I think we're ready to go to questions.

[Operator Instructions] Your first question comes from Adam Berlin with UBS.

Erik, I've got 3 questions, please. I was interested to hear you talking about the growth in Risk and particularly the growth in financial crime and compliance as being particularly strong. Can you just talk about why that the growth in financial crime and compliance is strong? It's not an area you talk about that often. So it would be good to get a bit more information about that.

The second question is about Legal and the Protégé product. What I'm trying to understand is, are you selling that Protégé product just to law firms who are not using Harvey? Or do you expect that law firms can take both Protégé and Harvey? And how do you see the difference between those 2 products? That would be really helpful.

And then just -- can you explain a little bit about what this next-generation AI Research solution product is? How is that different from what kind of AI-powered scopist does currently? And why do you think that's going to drive further revenues?

Yes, let me do those in order then. In Risk, yes, we continue to see very strong growth across market segments in Risk. FCC is one of them. Inside the Business Services segment, the 2 large segments there are, actually FCC, financial crime compliance, as well as fraud and identity solutions, those are the 2 big blocks.

And for the last year or 2, we've actually continued to see very strong growth in financial crime compliance. To a large extent, that's driven by the fact that, that we have, as we do across all our segments in Risk, an innovation machine where we keep driving new additional higher value-add solutions, leveraging new technology that we kept launching -- that we keep launching and rolling out. But in addition, this is a long-term structural growth market, and I don't see that the long-term structural growth in this is going to slow anytime soon. So it's the combination that we're seeing there of our product innovation, launch rollout machinery and the long-term structural growth of that market, which is very parallel to most subsegments in our Risk division.

The second question was about Legal. Well, we look at it this way, that our Lexis+ AI platform is the core integrated and we believe a leading legal research platform leveraging extractive AI and generative AI. And on that platform, we have built Protégé, which is the next generation virtual legal assistant effectively that gets to know you, follows you and helps you integrate any content from anywhere or any search tool or any type of other technology tool from anywhere. And our primary objective is to ensure that our AI solutions are accessible to our customers at every relevant touch point in their ecosystem. And as we do that, we collaborate with partners to integrate our AI offerings into all those platforms, including Microsoft and including Harvey that you said.

The way we look at it is that the legal tech ecosystem, which is much bigger than the legal research ecosystem that we're a part of traditionally, the legal tech ecosystem software tools and workflow tools there is a few multiples larger, and there are about 3,000 different vendors that we have identified there. And a typical law firm probably uses about 100 or 200 legal technology tools at some point. And our objective is to be integrated with as many of those as possible, and we currently maintain what we consider formal partnership with about 25 of those players in that ecosystem.

And therefore, we want Lexis+ AI and Protégé to be accessible at as many of those as possible, Harvey is one of those. It does not include or exclude any other sort of selling or any other approach other than what we're doing with every other partner that we have. We want our tools to be accessible at all of the touch points where our customers actually could benefit from that and see higher value.

The third point in STM, we launched in customer preview a few weeks ago, the -- what we refer to as our next-generation end-to-end AI-powered researcher solution. It aims to transform the researcher workflow really. And we think of it internally as the next generation of ScienceDirect AI and conceptually equivalent to what we've done in legal with Lexis+ AI and the virtual research assistant, conceptually similar to Protégé.

So the approach we're using, the methodology, the approach we're using in STM for the researcher solution is very similar to what we've done in legal with Lexis+ AI and Protégé. So that's the way we should see it. And we think that this is going to have a significant impact over time in the researcher world as well.

So just 1 follow-up, Erik, if I could just so understand it clearly. You could have a law firm that's using Harvey and Microsoft and all these different tools you mentioned, but it's also paying you not just for the Lexis+ AI platform, but also for -- additionally for the Protégé products, and that law firm could be paying all those fees, and that would make -- be a perfect strategy for them.

Yes, absolutely.

Your next question comes from George Webb with Morgan Stanley.

Erik, just a follow-up with regards to the new STM solutions you were just discussing. As you, I think, just mentioned, I think it's currently in closed beta, the Elsevier press release says available for purchase early next year that included if you already take ScienceDirect AI. Is that the right way to think about the timing of the launch and the monetization approach to outside of ScienceDirect to effectively be a new sale to customers?

Yes. Yes. The timeline, I have nothing to add on the timeline other than their press release because, of course, they communicate with their customers from the STM division first, and we just align with that from the corporate perspective. So the time line that they have announced is, therefore, the current time line with our customers. And it's true. You should think about it as a next-generation upgrade to ScienceDirect AI and that it's an integrated research solution that will then be sold separately, priced separately and with agentic capability that can be a virtual researcher system then built on top and several more iterations to come over time, just like what you've seen in Legal over the last -- well, we're really in Legal 2.5 years into now having continued to upgrade and evolve the Lexis+ AI and Protégé experience. I think you should think of it as conceptually very similar, but of course, now starting a bit later, but because of the collaboration and the coordination, it's going to follow very closely behind.

Your next question comes from Nick Dempsey with Barclays.

Erik, I've got 3, please. So first of all, you've called out double-digit growth for law firms and corporate legal within your Legal division. Based on the bookings that you're seeing currently from that customer base, is it fair to say that you would expect to see some acceleration in that line over time, so a slightly better flavor of double digit based on the bookings you are seeing?

Second question, just coming back to the new end-to-end research AI offering, presumably, it's a little bit harder to squeeze budget for incremental products out of this customer base compared to in legal, so can you just talk about the dynamics of how you are persuading those end customers that this enhancement to workflow is worth spending extra on?

And the third question, maybe you can just give a feeling for whether your renewal conversations with U.S. institutions in STM have been any difference to normal as a result of the potential challenges to their budgets that we've all been tracking?

Well, first, the double-digit growth in law terms and corporate legal, to those of you, many on this call have followed us closely over the years, it's probably -- it was probably already clear that if the division grew 9%, which we had at the half year, that the faster growing segment, law firms and corporate legal would be double digit and that the -- because the rest, the news and business and academic and government, has typically been a little bit of a slower grower. And so we described that as good growth. So, therefore, the double digit is more of a clarification of the 2 different segments, which I think many of you had already figured out.

But your question is actually what is happening to this going forward. The way we look at it is that the objective for legal is, as we always said, to continue on the improving growth trajectory, to continue to add more value to our customers, to continue to launch products that add more value and that they will adopt those, and therefore, use them more, use more of them, and therefore, spend more with us so that our growth rate continues to improve.

All indicators that we have is that we're on the right path on that in Legal as well as in other areas. And in law firms and corporate legal, we believe that we're on the right path to continue to do that. However, you do have to keep in mind that 85% of the Legal division is subscription-based, and it's typically 3-year subscriptions. And therefore, that the improvement in growth rate that you have seen will continue to come through gradually. And exactly when each piece will sort of tick over to the next percent on a rounded basis, it's not clear at this point, but we -- our objective is to continue to improve that underlying growth rate. We are on the right track to do that, but not clear exactly when it will tick over again.

The second question on researcher AI solution, just like with our other higher value-add analytics and decision tools that we've launched in Risk over the last 20 years, so we continue to launch in Legal with legal analytics over the last decade. And now, with the GenAI-based tools, we do not think of it as persuading our customers or trying to figure out if it's -- if they can squeeze more budget. We look at them as value-enhancing tools. And when we present them to our customers, they will see them and decide how much additional value they get from them.

And typically, the way they are priced, the way they are positioned, the way they add value is that we only price them to take a small fraction of the significant upside value that they see from them. So we expect that the customers that see the most value will be most eager to use it first. And then over time, as the product evolves and as our customers evolve and see the transparency of the additional value, we see increased usage, increased user penetration, and therefore, increased spend and gradually improving growth from that area. But we're seeing it -- see it as a value uplift to our customers, and we expect that they will see it the same way.

The renewals, it's too early to tell if there is any specific situation that you talked about in the U.S. at this moment. But I think it's important to keep in mind that this is a very, very global business for us. We have, I think, close to 15,000 institutional customers around the world. We have customers in 180 countries, and the different budget cycles go in different cycles at different times in different countries or in different scientific disciplines. We have been through many difficult periods over the last 200 years when we've operated our primary research business. And we always try to work with each one of our customers in each one -- each geography, in each situation to make sure that they can work their way through any challenges that they might have and that includes next year.

But as you might know -- also have seen that we've almost never seen those economic cycles have any material impact on the actual underlying growth of our STM division because it's so global, it's so diverse and it's so subscription-based. So we haven't seen any at this point specifically. That does not mean that there won't be customers that have significant challenges. But historically, when we work our way through this, it has not turned out to be material to the growth trajectory of that division.

Your next question comes from Sami Kassab with BNP Paribas.

Now, out of the 200 or so questions I had a chance to ask you over the last 20 years, these may not be the very best ones I'm going to ask, but this will clearly be my last. So Erik, Risk is growing at 8%, Exhibitions is growing at 8%, Legal is growing at more than 8%, so do you see any structural reasons why over time STM may or may not reach 8% as well?

Secondly, there has been a lot of talk on AI as a growth accelerator or as a growth disruptor, but can you please elaborate on how applying AI internally to your content creation process, to your customer and marketing service, to your own technology layer, how is that changing the nature of the business? How much productivity gains do you think generative AI can create for RELX?

And lastly, can you please comment on the copyright regime of open access articles? Can Elsevier and the other publishers decide on the copyright regime they want to apply to open access so as to prevent new entrants like OpenEvidence from accessing your content by applying the right copyright regime, say, CC BY-NC-ND? Or is the copyright regime on open access imposed on the publishing industry by the research funders, by academia, and therefore, not in your control?

Yes. Sami, so I'll take the first question, and then, I'll let Nick handle the second and the third question. Then I'll come back and cover the fourth question.

So on the growth rates of our different areas, as you correctly pointed out, Risk, Exhibitions, Legal, all now up at sort of 8% plus, which we consider very strong growth, STM in the long run should have the potential to move up towards those levels if we continue to leverage those tools into areas where our customers can see the significant value uplift that the tools provide them with in academic research, but also in applied research, in more commercially oriented uses of scientific research.

Research is a main driver of global economic growth. It has always been that way, will continue to be that way. And I believe that AI-driven digital tools that are content based and based on the vast content sets that we have will continue to help add value to the customers in the research -- global research ecosystem and as well in the global healthcare ecosystem. So I think the potential is there, and I think the potential is going to be there for a very long time.

However, because of how serious and regulated and sophisticated these industries are and how complex and complicated the underlying datasets are and the longer sales cycles in our customers, it will come through very gradually. So it might take us several years to get there, significantly longer than it's taken in our other subsegments. But I think the potential is there, and we can see those trends starting to emerge.

Nick, would you like to cover the next 2?

Yes. Sami, you're absolutely right. Of course, the AI and generative AI in particular is a big opportunity internally. Clearly, the most important thing about it is what it can do for our products. And we've obviously talked a lot about that already today. But also, as your question implies, we have the opportunity to use GenAI in our internal processes. And obviously, there are a lot of things we do processing content or coding -- software coding or use in sales and marketing or use in support functions, where process could be made more efficient.

We can use that to actually make our products even better, and you can actually process more content and have more content that you've been able to handle and tag and link and so on, which can make the product even better. So that's one use of using that efficiency to do things at the scale you couldn't before. And also, of course, you can reduce cost of doing things right across the board using GenAI. And that, of course, is what enables us, notwithstanding the resource that we're putting into new product introductions, entering adjacent markets and so on all the time that will cost money. But we can do that and still keep cost growth below revenue growth, as we've been doing for many years in all of our businesses. And that's why we believe using tools like generative AI to help us why we believe we can continue to do that.

And your last question on open access and copyrights, there's actually a range of different arrangements around open access and copyright. The vast majority of our open access articles are actually subject to copyright, but there is a choice -- customers have a choice depending on what they want to do, what their funding body requirements are. We offer journals right with a spectrum of different choices around copyright, and that's what we've always done.

Right. Then Sami, back to the fourth question, which you didn't ask, which is I would just like to comment on your 20 years, and I'd like to say a big thank you for having been with us for so long and so diligent. You've always been very knowledgeable, very inquisitive on these calls, but always with a constructive tone and your unique personal charm and friendliness. So I'd like to give you a big thank you. And also wish you good luck in the future.

Your next question comes from Steve Liechti with Deutsche Numis.

Can I please take 3 as well? First of all, just on Events, just can you give us any kind of clarity on forward-looking trends there, just thinking about next year anything to call out?

Secondly, just on -- in the U.S., the NIH APC sort of, well, proposals review period that's now finished, just any views there in terms of the options that they gave on APCs. And if you can give any clues on your average APC or at least a range? And I just wonder within that, sorry, for the question, just whether that might stimulate other people to limit APCs around the world.

And then the third question is just on your general AI product, which is now kind of launched out there. Just give us a bit more information in terms of positioning that product? And how that fits in with your kind of portfolio of AI products more generally?

Okay. I'll ask Nick to cover the first, and then, I'll move on with the others.

Yes. Steve, obviously, Exhibitions, we have what we considered to be a higher value-add business than we had previously with a higher growth profile, the Events, the portfolio we now have, which we slimmed down and focused on the real growth opportunities. And with the digital on top of that, that's what's helping to make that business a faster-growing business than it was historically. Obviously, at any moment in time, it varies from sector -- between sectors, between geographies, between particular events, but there's nothing to call out at the moment. As you can see from the results, there's good momentum in the business.

On the second question, regarding the NIH policy changes and evolving policies, the way we see it is that we've operated in this business, again, as you know, some of our journals continuously for over 200 years. During that period, we've seen a lot of changes from many different funding bodies in many countries. And it will always continue to evolve. We see our position in this industry as the largest higher quality -- highest quality player among the large players. We think of ourselves as having higher quality content, better technology and lower effective price per unit, unit of value than the other major suppliers.

And we can position ourselves around any change, in any regulation, any policy from any area and look at how we can position ourselves to offer the kind of quality tiers and the kind of pricing tiers that each funder is looking for. As you know, we had around 3,000 journals today. We launch between 50 and 100 new ones each year. They're all at different quality points. They're all at different subsegments of science, and they're all at different pricing tiers, but they're all on our leading technology platform and distributed that way. So we will always keep adjusting.

I do not believe that any one of these announcements at any given point in time will alter that strategy because it will always keep evolving as it has for the last couple of hundred years, and we will continue to evolve with it based on our ongoing positioning with very high submission growth, very high-quality content and the ability to continue to evolve our journal portfolio and our technology with it.

The last question, if I understood it right, was on Protégé general AI and how that evolves within the product ecosystem. Is that correct?

Yes. Yes, yes. Exactly.

Yes. Yes. So the way we look at it is that this is a continuation of our primary objective in the Legal segment, which is to -- again, to ensure that our AI solutions are accessible to customers every relevant touch point within their ecosystem and that our tools can be used with any of their other legal -- operational legal ecosystem tools. So that means that you can now use Protégé for your firm specific content, for our research content, for your own personal content. And if you then want to go do something that's what does the general AI tool say about this, how does that look compared to this, you can do that at the same time, integrate it, separate it, however, you want to look at it, but have the overlay of our trusted content, our verifiable tools on top of it.

It's going to be very clear exactly what you're using, how you're using it, how trusted it can be and so on, but we can do it from within our ecosystem, and Protégé can follow you and help you wherever you would like them to help you. So it's a natural evolution, and we think an important natural evolution that we know already that our customers value and that, therefore, increases their usage of Protégé at the different touch points. So that's how it fits in our strategies.

Great. And so to be 100% clear, it doesn't open up any of your kind of walled garden content to general LLMs and stuff like that, it's effectively just using the LLMs as tool within Protégé?

Yes, exactly. It is -- that's absolutely correct. We are using today many different LLMs. At the moment, we are -- we continue to be technology agnostic and multi-model by design for LLMs, so inside our legal generative AI tools today, we use more than a dozen different LLMs today, and they're all under contract as firewalled and internal and built inside our unit for our content and firm content. This is also being able to use them on other broad web and open content at one of the flows, absolutely crystal clear inside, which is which and what you're looking at and how trusted it can be, but it follows the exact same principles as a sort of industrial-grade internal firewall content. It's not putting our content there in their environment, it is using their tools in our environment and without putting our content there without any access to that.

Your next question comes from Thymen Rundberg with ING.

First one is on renewal discussions in legal. I was wondering if you could elaborate a bit on that and how those renewal discussions are progressing with clients, especially in the light of the rapid adoption of Lexis+ AI and Protégé. And I was wondering if clients are referencing better AI solutions as well or whether they're expressing new expectations as part of these conversations.

And then the second question, we discussed now the new AI-powered solutions in Legal and more recently in STM. I was wondering if you can share more about the pipeline for new AI-powered products that you're currently working on across your divisions? Where do you see the biggest opportunities for now?

Yes, on the legal renewal side, as we operate in Legal on -- mostly on rolling 3-year renewals, they take place throughout the year, and the main renewals take place every 3 years. It's 85% of that whole division is multiyear agreements. We continue to see and be involved in those discussions all throughout the year, which is different from some other areas, where they're on an annual calendar basis. But here, we continue throughout the year. And therefore, the adoption, the uplift that we're seeing, that we have seen and continue to see are a good reflection of that.

As I think we started to tell you a while back is that we also do new sales, and the new sales we have had the vast majority of all our new sales are of Lexis+ AI or Protégé or some combination of our integrated generative AI offering. And on renewals, we have a majority or actually a pretty clear majority of our renewal revenue comes from Lexis+ AI or other sort of generative AI integrated platforms at this moment. And that has continued. We've crossed that half of the renewals being generative AI inclusive probably about a year, 1.5 years ago now, and it has continued to increase a bit since then. So the trends are upwards, and we continue to see more and more adoption and more conversion at the renewal point.

If you look at the adoption curve, Lexis+, which was the first integrated legal analytics platform, integrating extractive AI, which was launched 4, 5 years ago, that was a very high-value tool, and that continued on a certain adoption curve that means that after 4 years in, we're sort of at 80% or so revenue penetration. This one -- this time around with Lexis+ AI, the penetration curve had been similar, but a little faster. And because it is a high value add, and our customers see it, it's probably moving slightly faster than the last version, the Lexis+.

And we have always seen competitors in our marketplace. We think of ourselves in the legal industry as a technology-enabled challenger. We've been focused on higher, more cutting -- higher value, more cutting-edge technology, a little bit faster than the established players in that market, the established, older research providers. We continue to think of that as a priority for us, and we'll continue to drive that going forward. There has always been competition. It will always be competition, but I haven't seen anything that has changed recently in that.

You said pipeline new AI product in the other divisions or across divisions. We continue to see significant opportunity to leverage new technology across the company in all 4 of our divisions. We have an established machinery for new product identification, launch, piloting and then roll out across the whole Risk division that we've been doing for a very long time, very established. It's mostly extractive AI machine learning algorithm.

The Risk division is now over 90% machine-to-machine embedded calculations. That machinery is going to continue to operate. We think there's a tremendous upside there, in particular as generative AI technologies enable fraudsters to do things at a different scale and with more sophistication, the defenses that our customers will need to build is very significant, and we will help them with that. So we see a significant upside there with new AI tools from our perspective to add value to our customers there.

Legal, I think we've talked about. We're still at the beginning of the impact of generative AI on the upside value in the Legal research business, where we've been historically, but also the generative AI tools opens up many workflow opportunities for us to go into areas that for us are white spaces in other segments of legal tech and legal workflow, where we haven't historically played, but it's a very large opportunity today and will continue to get larger.

In STM, we covered the long-term opportunity that it's going to come gradually, it's a complex global, fragmented industry, both in research and in healthcare, but the opportunities set is significant and will come in over a longer period of time.

And last but not least, in Exhibitions, we continue to build data-driven digital tools that we continue to build, test and launch in different industry segments at different exhibitions. And what we continue to see is that most of the tools that we introduce add significant value to the customers of that event. We see significantly higher value add to the customers who use the tools than those who don't. And we will continue to test, launch and roll out many of those tools over several years, and we see significant upside in value creation for our customers there as well.

Your next question comes from Henry Hayden from Rothschild.

Three from me. Firstly, with Risk, comments from one of your competitors in auto insurance indicated that LexisNexis has been kind of flexing its scale. Would it be possible to get some incremental color on this from your end? Is that coming from pricing? Is it more aggressive bundling, faster pace of innovation? Or is there something else? And further, what's kind of incentivizing this competitive ramp-up?

Secondly, on legal, what are you seeing in terms of the financial state of the underlying industry? And how would you assess kind of the state of the sales cycle? Is there elevated budget capacity for new solutions? Just any color there would be very helpful.

And then finally, just on capital allocation. We've continued to see leverage to come down below the target range. Should we expect kind of a continuation of the strategy of bringing that back up and returning that in the form of buybacks? Or is there appetite for more M&A at this stage? If the latter is an area you're specifically looking for a deal in, then is there a possibility that, that could be larger in scale?

Maybe we'll do this here. Maybe I'll ask Nick to cover the first here because I'm not aware of anything...

No, no. Look, I think we are -- we have some -- a very strong position in the auto insurance market. We provide a lot of value, and we're continually innovating. We've been growing that business for 30-odd years and introducing new products constantly. We have an approach of adding value. We don't raise prices of the existing tools because we create more value by introducing new things, and that's something we've been doing for a long time, and that's absolutely continuing. And it's a competitive market, of course, but we have a very strong position in it.

Now, on the second question, it might sound disappeared a little bit in the middle there, so I'm not sure I fully understood what you were saying. Could you please repeat the second question?

Yes, I was just curious on kind of the financial state of the legal industry for law firms and how you're assessing the current state of the sales cycle, how much budget capacity is kind of coming into the market.

Okay. Yes. No. So there are lots of different indicators of how the legal industry is doing that we look at mostly third-party research purpose studies and things that are published, and we then check that against our own experience without trying to build our own model, but we check it. We take them all in and we check it.

What we are seeing at the moment is that the legal industry is in, what I would describe as, relatively good shape at the moment. Things seem to be going relatively well from all those indicators. And that matches what we are seeing in our sales cycle that our customers are running their business. They care a lot about what they should be caring about, which is their customers and their competitive environment, their own performance and the service they deliver to their customers.

But they're also very interested in how they can leverage the new tools that are coming to the market, that we are offering, and we keep launching and explaining to them how they can leverage those to get a competitive edge in their markets so that they can provide higher value to their customers and change their competitive positions or operate better. So it's a resected industry to our type of product launches and rollouts at the moment.

And then, on the third question, I'm going to ask Nick to cover that again.

Yes. Henry, you're right in the sense that at the end of last year, our leverage was 1.8x. So below the sort of 2 to 2.5x range we normally target. And of course, we did announce a bigger buyback this year, probably to reflect that. And the last leverage number we published at half year was 2.2x. So that was in the range.

Now, of course, our shareholder returns, buyback and dividends tend to be first half biased. So midyear leverage does tend on average to be a little bit higher than year-end, but we'll see where it comes in at the end of the year. And putting that in the context or M&A in that context, the most important thing to say is that our focus is on organic development. I think, as we've been talking about on this call, the opportunities to grow the business and to roll out new products and add new value to customers organically is our biggest opportunity, and that's what we're primarily focused on.

We will make acquisitions where we think they can enhance and accelerate the organic development, but they need to fit with that organic development. So there are things that can help us accelerate, and we'll continue to look at things and see what opportunities arise on the M&A front. But it is with that approach. The buyback, what we said every year is -- as I've just described for last year, or prior to this year, is then used effectively to balance the overall capital structure and to keep us around that -- in and around that range that you described for leverage.

This concludes our question-and-answer session. I would now like to turn the conference back over to Erik Engstrom for any closing remarks.

Well, thank you for joining us this morning for our trading update. And I look forward to talking to you again soon.