Lloyds Banking Group plc (LYG) Earnings Call Transcript & Summary

Good afternoon, everyone, and thank you for joining us today. Having previously covered our growth priorities across our business units, I'm delighted to welcome you to the final investor seminar in this strategic phase, focusing on digital and AI, a very topical subject. These capabilities cover the full breadth of the organization and act as a critical underpin to our strategy. It's an area that myself, the Board and the rest of the management team have been extremely focused on over the last 4 years. And I was proud to see yesterday that our excellent progress has been recognized in Euromoney's latest assessment of digital banks, where we ranked within the top 20 globally out of a sample of more than 300 banks. I'm excited to be able to share with you some of the actions we've taken since 2021 as well as giving you an insight as to how we're thinking about future opportunities. I'm joined today by Ron van Kemenade, our Chief Operating Officer. As in previous sessions, I'll provide a brief overview before handing over. As always, following the presentations, we'll have plenty of time for your Q&A. Let me start on Slide 4. We're going to cover a number of topics as part of today's presentation, but the key messages that I'd like you to take away are as follows: Firstly, we're a digital and AI leader today and have distinct competitive advantages that will provide the foundations for long-term leadership. Secondly, since 2021, we've taken decisive action to enhance our infrastructure and underlying capabilities to enable change. The ongoing progress here creates a platform for innovating at scale, both today and in the future. Thirdly, as a result of this change, we're now delivering leading propositions and extending our digital capabilities right across the group for the benefit of both customers and colleagues. Fourthly, we're extending our leadership position into new and emerging technologies, and we feel well placed to succeed in a period of potentially transformational change for the industry. Taken together, these actions are delivering significant value for the group today, both in terms of driving revenue growth and improving efficiency. We see scope for this to accelerate meaningfully in future periods. On Slide 5, I'll provide an overview of our strong foundations. With 23 million digitally active users and over 21 million of these also using our mobile app, we are the largest fintech in the U.K, but we're also much more than that. Our scale drives unrivaled engagement with nearly 7 billion annual log-ons, it's also helped to create a unique data asset with the group processing 1 in 4 card transactions every year, equating to nearly GBP 330 billion of annual spend. Given we have relationships with over 50% of the U.K. adult population and more than 1 million businesses, nobody is better placed within financial services to understand their needs. As you'll hear shortly, we're unlocking more value from our data today and believe this will be one of growing importance into the future. On top of this, we have long since recognized that we must invest to be at the forefront of change, building strong AI foundations that will equip us with the necessary capabilities to continue to innovate at pace and deliver for our customers. Whilst others may be able to demonstrate strength across some of these areas, it is the combination of scale, engagement, data and capability that we believe provides us with a distinctive competitive advantage, reinforced by our strategic priorities. Turning now then to our strategy on Slide 6. When we presented our strategy in February 2022, we acknowledge that despite strong foundations, there was a significant opportunity to unlock the group's full potential. From a digital perspective, we had scale. But like all large financial services firms, we had a complex technology estate in need of modernization which limited opportunities to truly innovate and deliver at pace for the benefit of our customers. Our strategic priorities have been focused on addressing these challenges, and we've made great progress. We've further extended our scale leadership position, but have shifted the digital channel from one of purely servicing to one of engagement team. This has been supported by modernizing our technology estate in the areas that are most important to effectively compete. This means we can significantly increase the pace of innovation and extract more value from the unique data asset, unlocking new growth opportunities for the group. As our legacy becomes less of a headwind, we have also been able to shift our investment toward building future capability, new and emerging technologies such as agentic AI and digital assets, have the potential to significantly shape the future of our industry, reimagining customer outcomes. We're taking proactive action to be at the forefront of this change, cementing our status as the U.K.'s digital and AI leader. On Slide 7, I'll highlight how digital and AI leadership delivers clear financial benefits. The actions that we've taken during this strategic cycle are creating significant value for shareholders today. Innovative propositions and enhanced experiences are deepening customer relationships and driving revenue growth through both existing and new streams. In addition to supporting our ongoing franchise growth, the benefits of our investment in specific digital and AI initiatives will contribute more than 70% to our strategic revenues target in 2026. At the same time, we're driving efficiency gains across the group, lowering cost to serve and delivering a more agile organization with a more flexible cost base. More than 60% of the circa GBP 1.5 billion of cost savings realized to date are directly attributable to our digital and AI initiatives. This is great progress and a sizable contribution, but encouragingly, we see additional upside beyond the current plan. Our commitment to investing in new and emerging technologies has the potential to unlock further value, improving operating leverage through new revenue growth opportunities, drastically reduced acquisition and servicing costs and a step change in productivity. We'll discuss some of these areas today, and you will, of course, hear more about these benefits when we update you on our strategy in 2026. Before handing over, I'll briefly introduce you to Ron on Slide 8. To deliver such a significant transformation, it was critical to put in place the right leadership. I was, therefore, delighted that Ron agreed to join the group in 2023, bringing a proven track record of leading major technology transformations. Ron has a clear understanding of what makes companies best-in-class and is able to keep us honest throughout the journey as you'll hear, as well as providing the necessary expertise and the leadership to drive the change. Ron has a broad remit at the group, bringing together multiple areas to deliver a technology center of excellence. His teams are a key enabler to our strategic growth priorities with connectivity to our customer-facing business units established through integrated CIOs to help foster innovation. I'll return briefly at the end to explain why I'm so excited about the opportunity that new and emerging technologies present for the group. But for now, let me hand over to Ron to talk you through the journey we're on. Ron?

Thank you, Charlie. Well, after such an introduction, I probably can only disappoint you. Let me talk you through indeed our digital and AI strategy. And if you can't pay attention for all the 12 slides, just remember 3 things. When I joined, like Charlie said, mid-2023, neither Charlie, nor Robin Budenberg, made it a secret that there was a lot of work to do, addressing way of working people, technology, the legacy estate. But at the same time, they did convince me with the opportunity, because it's an honor to be the technology data, the group COO leader of a bank that is largest financial institution in the U.K., having impacts, like Charlie said, on half of the households, 30% of every payment processed in this country. There is a huge opportunity to become the biggest fintech of the U.K. And that convinced me, right? And now 2.5 years into the journey, I can confidently say that we are effectively addressing our past that we are delivering on our strategic objectives of the growth focus and change strategy until '26. And then we are laying a solid foundation for innovation and base in the future. And let me elaborate on those things in the succeeding slides. Let's start at the beginning. There was obviously a legacy landscape, all banks have legacy, we as well. It's a matter of how you address it, right? And dealing with modernization of your infrastructure is critical to lay a foundation for pace, for cost effectiveness and for innovation for the future. And we are making good progress. On the lowest layer of the stack, we are consolidating our data center footprint from 18 in '21, now down to 9. And by the end of '27, we'll reside most of our tech landscape in 2 hyper modern scalable data centers next to our public cloud usage. Secondly, if you look at our application landscape, we are simplifying hugely, having taken out 20% of all the applications by pure decommissioning. And at the same time, of the remaining landscape, we have modernized half of it, bringing us to a point where legacy is not in the way of pace nor innovation anymore. And you could say, well, Ron, there is still 40% left, and you're right, but we consciously chose 3 major platforms that we believe are crucial to deliver more value at pace for our customers and colleagues. First one being our digital platform, enabling our digital channels. Secondly, our data platform, so we can unlock the value, as Charlie said, of our rich wealth of data. And thirdly, our core banking platform, which unlocks new products for our customers. Second part of our transformation is, of course, about people. And arguably, this is the most important part, one I'm extremely passionate about. We have decided, we want to build a workforce of highly engaged, highly motivated, purpose-driven and above all, highly skilled engineers in data, in cybersecurity, in AI and in tech. And in doing so, we are recognized in the market. We are able to attract top talent from recognized tech companies from Amazon, from Google, and you've seen the announcements of recent appointments of our Head of AI, our CIO for the commercial bank as good examples. Secondly, we have hired over 900 graduates, people coming in fresh from university with a new mindset who are, I would say, born with technology to whom a mobile app or AI is just something that they have grown up with. And we have hired over 8,000 engineers over the past 4 years, effectively changing the ratio from 30% permanent workers to 60% on our way to make it 80% by the end of next year. Again, a workforce that is committed by signing up as an employee for Lloyds, people who know our purpose, who live the experience of our services every day, either working in the U.K. or in India. Third part of our transformation is the way we work, and we have effectively moved all of our change and run activities into a platform model. And you've heard words like platform, maybe you've heard of words like Tribes my previous employer, it's one and the same, right? We combine people from technology and business into one organization that fully own, run and change that are joined up by one backlog that is fully product-driven to deliver value that share the same objectives where the business is leading on what we are supposed to do to deliver on our strategy and the tech people worry about how to deliver this. And a very good example, what benefits this delivers, this platform model, is the recent introduction of a new personal current account for our mass affluent customers called the Premier accounts. And arguably, a personal current account is one of the most complex products to introduce in a bank. It is connected to literally everything to the lending domain for overdrafts to payments, obviously, to the customer well-being or arrears management platform, your channels, both colleague channels as well as digital, it ships data to marketing, risk and finance. So it's probably the most connected product. And we delivered a fully new proposition just within 5 months, where if we look back to previous introductions, this would have taken us arguably in between 12 and 18 months. And I think 12 is on the more optimistic side of the spectrum. And next to the pace, the platform model does help us reduce the cost to deliver. In gross savings, we have saved over GBP 300 million annually so far, which if reinvested, adds only to additional capacity to further accelerate our change. Now this is the effort, right? This is what we have done. The question is, what does that actually deliver for our customers? And let's look at our mobile app, and Charlie already alluded to this, highly recognized by Euromoney in the benchmark. I think we're up in the right quadrant at somewhere #5, 6 or 7, close to ING, by the way, that's no coincidence. Yes, it's slightly ahead, Charlie. That's what you get. Indeed. So it is arguably our most important channel. So that's why it's so relevant to look at the impact there. In terms of engagement, we are now driving over 13 million fully personalized engagements a day. Just imagine that by sending out 20 million notifications that customers react to. In terms of sales, the channel share for sales has gone up 20 basis points from 55% in '21 to now over 75% of all of our product sales go through the mobile app, which is a huge impact. And in terms of our impact for customers, we are now embedding AI, which makes their interaction with the mobile app even more engaging, it feels more natural because you can simply talk, and it feels you talk to a person that gives you relevant answers, which is arguably a more intuitive and natural way of interaction than just clicking on buttons, going through forms and dialogues. Now let me give you one example that I'm really proud of. In only 8 weeks or we express this in sprints of 2 weeks, but let's say, for normal people, 8 weeks. We fully redefined our onboarding for new customers. So from downloading the app to having a personal account that is actually reachable, you can fund it and you can withdraw money, you have your card in your wallet, your virtual card in literally 7 minutes, just imagine, right, if you would start this onboarding right now, which I'm sure you will all do, you're eager to open an account with Lloyds, I'm sure. You would be done before I'm actually done with my presentation. And this is on par level with our competitors in the market, in particular, the neo banks. And the impact of this, again, for the bank, the value is huge, where only 2 years ago, 20% of all the personal current accounts openings were done through the mobile app that has increased to 90% today and still counting. So where we talk about the impact on just one channel, the mobile app, the question obviously is, do all these investments actually scale? Is there leverage to it? And the answer is yes. And let me talk you through this in 3 kind of examples. The first one is that digital platform that I was alluding to, which is the basis of the successful mobile app. We are reusing this across the group for BCB for IP&I in the Scottish Widows app, which is built on the same digital platform now scaling from 0.7 million customers this year to 1.5 million, which is our ambition next year, again, using the same platform. Second example is our business current account. I just told you about the beautiful story of onboarding in literally minutes, in the business banking account, this has reduced the time to onboard from an amazing 30 days to now 2 days. And 2 days is still not where we want to be, but it's just an example of reusing the same investments that we did for consumers is paying off in the business banking domain as well. Another example is in the domain of AI where relationship managers for real estate finance that need to capture these very complex tenancy schemes that they need to kind of take from all kind of documents, which normally takes them hours to do, it's tedious work and it's administrative work, obviously, distracting them from interacting with customers. By implementing our generic AI capabilities has brought that down to less than 5 minutes. Just imagine what that unlocks in terms of value for our relationship managers in interacting with our customers. Second way in which we are growing our business based on all the investments is how we have introduced ecosystems, something I'm really proud of because there is this question of how do you engage customers that initially apply for a product, you prove them and then what, right? And in particular, for mortgage customers, this is an issue because customers do not have particular reasons to frequently interact with the bank unless they want to repay or they want to remortgage or whatever. But in between, there is hardly any engagement. And I think we really cracked this equation by introducing our Homes Hub, where we now have over 450,000 new reengaged interactions where customers are benefiting from insights about homes and living where we have offered additional services serving broader customer needs. And in doing so, we have retained GBP 10 billion of mortgages in just over 1.5 years. And GBP 10 billion, as you will all appreciate, is more than many mortgage providers actually have on their books and only this in 1.5 years. And having built this ecosystem capability, we are now leveraging this across to transport. One of our most important businesses on the lending side, where we are now offering additional car insurance products and more mobility services contracting with other partners in our ecosystem. So a good second example how initial investments pay out across the group. And then to our data, as Charlie already said, we are, you could say, figuratively sitting on a wealth of data. We know more about British households and companies than any other company in the U.K. It's a matter of how we unlock that value and grow our revenues by doing so. And we have enabled this by investing in our new fully cloud-based data platform, unlocking AI capabilities and analytics capabilities. Two examples here. Personalized offerings. And a good example, again, here is from our mobile app for consumers where your credit score has been introduced, where customers can see what their actual credit score is as it is registered what components of that are, and we offer advice how you could improve your credit rating. It's fully personalized. Again, 12 million customers to Charlie's point of scale, 12 million customers have registered for it and are using it. And it now accounts and this is only the start for 10% of all lending leads. Second, example of how we are using our data to drive revenues. We are offering data products directly to our BCB and CIB customers in the commercial market, offering products under the name of Market Intelligence, where we are able to tell customers based on obviously anonymized data what their market development and market share developments are actually compared to competition, which drives a huge value for those customers, driving deeper engagement with those customers, and helping us to ancillary your business. Now all of this is kind of addressing my first point. So what have we done to address the past and do we deliver on our promises today? And as you can see, we believe that we are effectively doing this. Legacy is not in our way anymore. We are leveraging value out of our initial investments across the group and we have laid a foundation for innovation for the future. And let me address this as the last part of my story. What about the future? There are 2 major technologies that we believe will define the next couple of years. One is generative or agentic AI, no surprise to you. And the second one is digital assets, and I'll talk about them both. We have invested into a generative AI platform, making use of the best available capabilities in the market open to different LLMs in the back, using our data in a safe way with guardrails around it to make sure that we have ethical unbiased propositions for our customers, and we have moved beyond and this is an important message, the phase of beautiful experiments, promising proof of concepts and small pilots. Every investment we make needs to go to production. So we have committed to deliver 50 use cases into production this year alone, and we're on our way to meet that target bringing GBP 50 million of concrete money in the bank benefits this year only. And that's just the start of it. This will scale substantially, and I'm sure, Charlie will come back to this next year in our strategy update. This will scale substantially into the future. But that 50 million, and I can't emphasize that enough is real money in the bank. It's no future promise. It's validated by our people in finance. So we do invest in kind of 5 areas, like I said, directly in our customer interaction in our digital channels. Secondly, supporting our customer operations people, our relationship managers, and I gave you the example of real estate. We offer generic HR support through a modern proposition called Prosper to all of our 60,000 -- sorry, colleagues. And we are investing in improving productivity for our engineers making use of GitHub and other tools. So those 5 areas. And I'm glad to say we are, again, seen by the market as leaders in this domain. One is Microsoft, who has called us out as one of the leading financial institutions in adopting AI, and if you don't believe then, look at the survey of evident AI, which ranks financial institutions globally, where we were #27 last year, now we're ranked #15 and moving up, right, which puts us right at the top of U.K. banks, which is obviously relevant to compare ourselves to. Now these are cases that I'm sure you're more or less familiar with, but this is not where it stops. We are using generative AI and in particular, agentic AI to reimagine our propositions and there is no limit to ambition here. As you all know, one of the big unsolved equations in financial services is how do you scale financial advice whether that's for pensions, investments or any advice intensive products without adding people. How do you scale this in a digital way, and we are doing this. It's a proposition that in the consumer market actually has been introduced yesterday as a pilot for friends and family. It's called CoachAI, which is more like a broader adviser. In the next couple of months, we will introduce a similar proposition specifically for investment advice called InvestAI, and we are proud that we have been elected one of the just 5 companies to be allowed in the experimentation and innovation environment of the FCA. So they can work with us that we are working on a proposition that actually meets the regulatory requirements as well as meeting customer expectations. And let me show a short demo of this to be launched proposition. [Presentation]

So big ambition here for how we want to use GenAI and agentic AI to reimagine our propositions for our customers. Let me finally move on to the second area of innovation, technology-driven innovation that I announced and that is digital assets or and/or programmable money. Two examples here. Again, we believe we are well positioned as a -- not just a thought leader, but a very early mover, and in doing so, we are positioning ourselves as a leader in digital assets as well. We have started a new lab and platform for this. And we are co-sharing, actually, the U.K. finance platform for digital assets and so-called GBTD, which is Great Britain Tokenized Deposits. And for those who think Jesus, what are tokenized deposits, that's a fair question, right? Let's say, a tokenized deposit is nothing more than a deposit, as you know it, so sterling money, but then deployed as code as smart contracts on a blockchain. And why is that so relevant? Why do we need to digitize money because it gives us the opportunity to put logic to the transaction. For example, in peer-to-peer payments, you can add conditions to it. How do we transfer the money only when the service has been delivered, right, which a good example, and I'm sure Charlie will say a bit more about it, but in the conveyance of mortgages where you need a whole system of solicitors, notary services to actually make sure that there is trust in the system when one customer hands over the money, the other customer hands over the deed to the property. Digitization of money gives you the opportunity to do exactly that with all -- without all of the intermediary steps, because you can program that condition into the money that it will only be released at exactly the same moment, the digital deed is being transferred. I think that's a great example of how digital money, in particular, tokenized deposits can actually help unlock a lot of value for the larger U.K. economy. Second example is we did a U.K. first transaction with Aberdeen Asset Management, where we did an FX forward, I think, for 2 months. And then in -- as you're undoubtedly familiar with, in an FX forward, you always have the margin goals, right, if the actual price is deviating from the set price. And to settle those margin goals is a hassle. You need to provide collateral if the margin call is in your favor, you need somebody to repay you and the other way around. There is transaction cost involved, you need to sell the collateral, you need to cash it, then transfer the cash. And in tokenization, in their case, of a money market fund and on our side, a digital bond, a digital gilt, we have been able to make margin calls real time, so you don't need to sell cash out and pay, but you can simply transfer part of the tokens of that digital asset, which lowers transaction cost for margin goals for the customers, and it speeds up the process from 1 to 2 days to actually instantaneously. So real benefits to customers and real benefits to us. That's why we believe digital assets are such a relevant second area of innovation in the market. Third element is we are actively working with government, with the consortium of banks to work on digital identity and digital verification services, because, again, we are sitting on a wealth of data, and we are best positioned to verify people's age, for example, of our people's address or even people's affordability of certain things. So again, the third area where we believe there is, again, we are only at the start of a new ecosystem of services where we are in front seat. And with that, Charlie, I hand back to you.

Thank you, Ron. So Ron has just highlighted some of the exciting things we're doing today. But before closing, it's worth briefly stepping back to explain why we believe new and emerging technologies will present such a transformational opportunity. Some of the examples Ron was just talking about. Technological change has been a constant over the past 3 decades and has transformed the way in which the financial services sector operates. Looking ahead, as agentic AI and digital assets become increasingly more mainstream, these present the next likely step change for the industry. Like with all change, this will present both opportunities and threats. However, given our existing leadership positions, combined with broad-based support amongst government, Bank of England and the other industry participants to position the U.K. as a leader for these new technologies, we see more opportunities than threats as we look out today. As an active participant in these discussions, we believe we are well placed to shape a future that drives positive outcomes for our customers whilst at the same time, unlocking profitable growth opportunities for the group and our shareholders. I'll explain for an example on Slide 25. Agentic AI and digital assets have the potential to be highly complementary with customers interacting with agents as a channel and digital assets, increasing the ease of transaction execution. For example, it's possible to imagine a scenario where home buying experiences could be enhanced by agents, scanning for the best deals and undertaking the transaction on behalf of customers with tokenized records and programmable money, ensuring a seamless conveyancing process. For our customers, this has the potential to drive a meaningful improvement in experience with reduced effort, greater security and real-time settlement. It also presents opportunities for the group, including lowering costs to both acquire and serve and increasing direct engagement with customers and what is a highly intermediated market today. It's clear to me what is required to be a leading provider in that future. Firstly, trust established over many years will be hugely important to both customers and regulators. Secondly, a broad offering will be necessary to maintain relevance and remain front of mind. Thirdly, those with the deepest pools of data will be best placed to train models on financial interactions. And finally, targeted investments in capabilities will be critical for leveraging AI at pace. This group, Lloyds Banking Group is uniquely placed across all 4 of these areas within the U.K. Our scale and breadth of businesses spanning retail and commercial banking, insurance and wealth is unrivaled, and our unique data asset will become of even greater importance than it is today, benefiting from the infrastructure investments that Ron just described in this strategic phase. Combined, we have a truly differentiated position compared to all other providers. We're going to talk more about these areas more in detail when we provide the next phase of our strategy in 2026, but I'm highly confident that the actions we're taking today will continue to position the group as a leader going forward. I'll now close on Slide 26. In summary, we're building upon unique competitive advantages to reinforce our position as a digital and AI leader. We've made significant progress since launching our strategy in 2022. This is driving significant value for our shareholders today, contributing more than 70% of our strategic initiatives revenues by 2026, and more than 60% of the gross cost savings realized to date. Our strong execution to this point means we are well positioned to take advantage of future opportunities, extending our leadership position across new and emerging technologies. This will further enhance operating leverage through new growth opportunities and continued improvements in efficiency. Thanks for listening. I'll now hand over to Douglas who will facilitate the Q&A. Douglas?

Thank you, Charlie, and thank you, Ron. We now have plenty of time allocated for today's Q&A session. We will start by taking questions from the room, but we'll also cover questions submitted online throughout the session. [Operator Instructions] Okay. So let's begin. I'll probably just start over that side. And just for fairness with the microphone lady, we'll run across. Guy, why don't we start with yourself.

It's Guy Stebbings at BNP Paribas. The first question was on ready-made investments, which I think is sort of a really exciting proposition. Could you tell us around the sort of expected time line for the full launch of that? And then working with the FCA in Sandbox, how should we think about that in terms of safeguards around any future claims around sort of quality of advice and that sort of thing? My second question is on the Homes Hub refinancing. Can I check that 15% definition? What that means? Is that sort of 15% of total balances of refi through the Home Hub? Is there other refis that happen outside of that, just kind of keen to understand the definition and how that's evolved over time. Clearly, that could be quite sort of powerful from a value perspective if we see more and more going through that channel?

Excellent. Thank you, Guy. Probably makes sense for you to kick off with those questions, I suspect, Charlie.

I think it probably does. But well, first of all, thanks -- thanks very much for the question. And what's great about these is run teed up. You can see we're trying to make sure what we're doing around digital AI is leading into better propositions and enhancing the areas. Just in terms of ready-made investments. We already have ready-made investment product line, as you know, and it's been part of the success we've had to date. I think I've mentioned in previous sessions, we've gone from 9% or less than 9% share of equity ices up to more than 20%, and we continue to trade at that level of growth in our market share, and it was partly based on our ready-made investments journeys as well as some of our self-directed channels. The exciting thing, as you say, is the industry actually in the whole world, they've introduced robo advice years ago, but it hasn't really had any intelligence in the advice journey. It's just been a structured logic journey to get through to simple rebalancing ETF broadly. And the journey that Ron gave you an example of a kind of conversational advice journey, which also uses the existing data the customer has to give them confidence to make the right choices and that will sometimes by the way, not be investing, it would be better to pay off existing, for example, credit card loans. That's the journey we've got in this regulatory sandbox. We're intending to get it live in the Q1 next year. So we'll be going through, as Ron described, we have a big user base internally with families and friends that we do co-development in a way that all the firms that you've spent time covering do who are not big banks. We've been doing that for a while now. So we'll be working on that basis in the next period of tough time and then going live with customers. And yes, the regulatory sandbox, we think is really important in this context. Obviously, whatever we do, we want to make sure is endorsed by the regulator. There's been a significant change around the regulation that underpins investment advice, this concept called targeted advice or targeted support, which means we already have a regulatory framework, which is much simpler than RDR, which is the regulation that underpins investment and resulted in this -- the majority of the market not having access to investment advice. And so we do have a regulatory framework that's clear around that. But as you rightly pointed out, Guy, actually getting the way the conversation on the journey works and having clarity with the regulator that they are comfortable and we are comfortable that gets the right outcomes is really important. So one thing I think coming out of this is when you see this land, it will be a fully functioning journey, but this is one of the things about agentic AI. It will be based on our data and our training of, in many cases, millions of interactions of customers that will improve the journey and the journey will improve over time. The other great thing about these journeys that we build and then the technology we now have is we can collect a lot more data around how customers are actually going through the journey where the fail points are and their understanding. And again, it's the engineering team working together with the experts in our investments and advice journey that will then use that data to make these journeys safer. And I say this with a smile having run wealth businesses for a lot of my time, instead of relying on 1% or 2% callbacks in sales quality. We can build agents, monitoring agents and do 100% quality checking, and that's the kind of mindset that we're going to be building into these journeys. So you should see it improve materially over time. The final thing to say is that, look, it should improve what we're capable of doing today, which is already delivering value for the group. But as we know, if we really start to -- this is not really a phrase by any way, democratize wealth, we get wealth and investments to be something that the mass market really can safely get access to. It's going to take some time for them to engage to decide to put money every month aside for investments and for those balances to build. So we think it's strategically really important in terms of relationship, trust. And over time, it will build a very attractive bluntly return business but it's not going to happen overnight. And then your second question around Home Hub, look, I'm going to give you what I think the answer is. But I think, Douglas, we should just make sure offline. I think that's relative to the whole of the refinancing market. As you know, we are, by some way, the largest mortgage provider in the U.K. What's interesting, of course, is, as you know, only 40% of homes have a mortgage and only -- the average LTV is about 50%. So when you look at the numbers of customers, actually, there's not that many people who are refinancing a mortgage, and we are by far the biggest. So getting to 15% is material. Ron mentioned the over 400,000 customers engaged. That may feel low relative to the GBP 28 million. But relative to the stock of mortgage customers, it's a very, very high level of engagement. And that's why this -- we are very excited about this Homes Hub tool, it really is material in terms of customers engaging with it and then us having an opportunity to build a broader relationship. The one other point about the Homes Hub when you look at it immediately is we happen to operate in this market with 85% of mortgages distributed through brokers. And so we find our customers who are broker originated may not have a relationship with the rest of the bank, engaging with this Home Hub looking at protection, thinking about our retrofit solutions, understanding the broader offers of the group. So it's actually one of these brilliant tools for helping ensure we give good outcomes to our very important mortgage customers, but engaging them on the group more broadly because the U.K. has got to this place where intermediation has become material, and as you start to think about the future and us connecting with customers and providing agentic AI advice, we're pretty excited about where we could go.

Let's take James behind.

It's James Invine here from Rothschild & Co Redburn. I've got 2, please. The first is that, I think in the presentation, everything was about Lloyds Banking Group products. I was just wondering if there's any scope to combine this with open banking agenda and help customers to manage their finances across a lot of different providers. And if you don't do it, do you think somebody else will? And then the second question is, it certainly seems like this is a great platform. Do you think it would be -- do you think it would work internationally. Have you considered opening a digital-only bank somewhere else? I know you wouldn't have all the customer data, but you would have the infrastructure?

Thank you, James. Ron, I suspect probably the open banking side will probably be worth you addressing to start with, and I suspect the international element will be Charlie.

Brilliant question actually because as Charlie alluded to this, GenAI is based on LLMs, right, which provides generic insights. It creates generic knowledge. And then you have your own view on the customer based on the product position today based on their channel behavior, based on what you know about their creditworthiness, all of this. Open banking definitely broadens that picture again where you add to the relevant context, knowing more about the customer than you do today. Now as you're aware, we are offering open banking APIs to the market as well as we offer our customers the opportunity to open their accounts and other products into our digital channels, and we are actively integrating this into CoachAI and further AI-based propositions.

Great. Yes. So it is really interesting. I don't want to take the question further, James, but my guess is you're thinking as we build intelligence that's really supporting customers, how do they start navigating the market. And so we are actually -- I don't know whether we are. My guess is we are -- I should know this data point. My guess is we're the biggest provider of integrated products across providers through our digital app because we've got by far the biggest digital app and many of our customers I can see my head of this over here. I'm not going to ask the answer to this, many of our customers already open show their other products and other banks in our app, and we see that data today. The really interesting question is what happens next is the intelligence in our app and then through other LLMs in the market start to provide advice to customers about who they're going to shop, how do you provide the best advice and how do you compete? And that's where, again, being at the forefront of this with the best data to get the best outcomes to build the trust, we think is the most important thing. And I'm not being vague and avoiding it, but I think there's an opportunity and a threat in that, that we feel were placed and positioned to look after. And look, just on the international side, one of the things that I say a lot to our team is we are staying focused on delivering the current strategy for now. And through 2026, we've got some pretty bold commitments that I know many of you and all of you are aware of and you've got real clarity that we as a team are committed to delivering. And at the starting point, when you look at the value of the group and the value coming out of our U.K. businesses, specifically, delivering this agenda, both in '25, '26 and then for the potential beyond is huge. The kind of cash flows, the outcomes and then as you know, our commitments to greater than 15% RoTE 200 basis points of capital and a cost-income ratio of less than 50% by next year is forefront of our mind. So that's where we're focused now. Definitely, there are options for us to do things internationally. We will be, and we have been having a discussion around that as a team, and we'll come back on the next phase of the strategy. And I think you teed it up well. We know we have one of the best digital platforms and Ron and the team have just gone through an excise of modernizing the infrastructure behind it. There's a bit of a way to go on that. So that could be a choice for us. But we always need to come back to what's our core strategy and what's the best return for our shareholders and we look at our option in that context. So no commitment at this stage. But I think when you become -- we're going to end up being one of the biggest and certainly most successful digital retail and then SME, we're slightly smaller, but digital retail banks in the world. We will have very distinctive capabilities that we could take elsewhere. But no commitments around that because I want to stay focused on the current strategy.

Just go to Alvaro.

And are you comfortable with that answer.

Alvaro Serrano from Morgan Stanley. I have 2 questions, please. First one is on the current account upgrade, which is kind of trying to tie that with the other income initiatives. The vision when Scottish Widows was revamped, my understanding was always that you would be able to access all the current account data and predict when there was an insurance payment coming and offer Scottish Widows an alternative product or any other parts of the product suite that Lloyds Group has to offer. Is this PCA upgrade sort of can you do that today? If so, how much sort of you alluded to uptick in revenues, I think, or I don't know how you phrase it exactly, but are we there today? What's the revenue opportunity or when do you see that happening? And then the second question is around tokenization of deposits, and you've mentioned and it's been widely reported sort of the efficiency and the convenience is clear on the lower cost obviously, Lloyds and the banks make money out of the payments. What risk are we seeing from a revenue perspective versus the opportunity? How should we think about sort of tokenization more broadly, not just deposits?

Thank you, Alvaro. Ron, it's probably -- I suspect, on the first bit there about the integration between insurance and banking, it probably makes sense for an initial -- from your side, and then we can move on to tokenization of deposits.

There is 2 opportunities, I think. One is we have these 22 million almost digitally engaged customers on the banking side, where we offer the full breadth and depth of all of our propositions, products and services, including the Scottish Widows products, right? And this is the, the famous bancassurance model. And increasingly, we are successful at that. On the other side, you have the Scottish Widows app that is unlocking more and more insights into people's products, whether that's their pensions products, protection products, P&C insurance. We are offering those customers the opportunity through open banking APIs to give insight into the banking products as well. Now how that exactly will materialize into additional revenues, either on the insurance or personal current accounts or other products on the bank side. That's a bit of a guess for the future. But obviously, the intention of the whole cross-fertilization across the Scottish Widows digital engagement on one side and the banking digital engagement on the other side is exactly that what you're hinting at. And maybe, Charlie, wants to add to that?

Well, one other add I think Alvaro, the question was, are we doing it today? And are we predicting for example, someone that doesn't have an insurance relationship with us, but we can see as a relationship with somebody else, and we should be putting in front of them offers of our product. And are we starting to think about providing preferential product and discounted pricing. The answer is yes, we are. It's not as advanced as we can see it could be, and way you've talked to you before about the increase in us bringing -- I'm going to use the word cross-sell, cross-selling bancassurance product to our core customer franchise. We've increased that very significantly on the back of mortgages on the back of home insurance and the Home Hub, the enhanced mortgage journey and then this personalized messaging, and Ron talked about the data earlier around how many messages we put out. That's been the progress that's helped us do that more successfully, more joined up journeys with better insights and then in the language of the nerds like me the contextual relevant message. It's contextual to the customer at a time that's relevant through a channel that they want to access that channel. That's the intelligence that we have a leading capability in the U.K. So we do all that, and it's been driving those improvements we've seen. When Ron and I sit together with our teams, and we talk about what's going to be feasible based on how we now have the data and then the use of agentic AI. And then as we move to a more conversational banking engagement through CoachAI and InvestmentAI, I think we're going to become incredibly better, much better at doing this. And so there's a real opportunity for us to really help customers get better outcomes with products that we can then price in a way that gives them more value across the group. So -- it's been a part of what we've delivered to date, but actually, there's a step change still in front of us that we are positioning for based on these new technologies. So I'll go to the second one. Just on tokenization and deposits. Look, this is a really broad question. So tell me whether this kind of covers it right. Look, there's both opportunity and threat. And to be clear on the way we think about GBTD, which I want to call brickcoin still, but nonetheless, no one else does. I'm told that's for 1990s. The -- this is largely around domestic payments. And interestingly, of course, we do have -- it's wholesale and retail. We do have some of those payments that have a fee, but the biggest platform by volume by number of payments is faster payments, which is actually free. So interesting, first of all, from a pure transactional payment perspective, there's a combination. And actually, some of the bigger use cases would be the faster payments mechanism. As you know, we process 20% to 30% of domestic lower-value payments every day. So we are right in the center of that activity. Look, we see it as both an opportunity to grow and also there's a defensive part to it when you think about the future of this. The opportunity to grow, as we just described is we're going to be able to build programmability and embed the payment in broader transactions in the way we said. And if we're better at doing that than other people, we can help people when they're on their online P2P purchases, whatever platform they're using in their mortgage journeys, businesses doing B2B transactions, trade transactions the integration of the digital asset programmability and then how we embed that in the broader journey for the customers, we think -- we do think will be a source of advantage, competitive advantage for those that do it better, and we'll be able to do that as a leader in deploying that technology, and that will give us growth. And in many of those products, the reason I went more broadly is it won't be the payment that brings the value it will be the broader transaction. And in some cases, there's fees and in some cases, it's around broadening and deepening our lending or deposit relationships. That's the first thing. The second thing is, and Ron gave a great example in the context of our money market fund and gilt digitization. We do think there's a significant opportunity for efficiency improvement, not just on the payment, interestingly, but also on the associated document sharing and then engagement of the customers and the fulfillment of the transaction. And we've -- many of us have been doing this for 30, 40 years now. The history of financial services through my whole career has been one where the application of technology has resulted in margin compression, and you need to then scale, be more relevant, do more for your customers and get efficiency. So we just see when you look out into the future, this is another one of those opportunities to build efficiency. So those are both opportunities for us and our shareholders. The defensive side to it is, especially on the back of the stablecoin announcement in the U.S. and the GENIUS Act and the STABLE Act and the derivatives they're coming, is if we saw a significant shift on domestic payments for -- to any kind of stablecoin, you then get the question around, does liquidity start to shift out of the banking system. And so our intent here is to very rapidly because we think we've already got pilots live, and we think we can get this live for the whole system, the whole banking system by H1 2027, Q1 '27, subject to working with our regulators, which would be the first market in the world that's done this. Singapore, China all my favorite ones I always beat us up on, haven't done this. The opportunity is, I think, for -- to make tokenized commercial bank deposits as good for our customers and better from a programmability perspective than alternatives. I clearly think they should be interoperable with international stable coins at some stage, and we'll work with our regulators on that, but there's a defensive play around that, and we think it's really important to make it a better customer experience and more integrated, and that's one of the reasons we think it's a really important thing to do going forward. So there is a defensive part to it, but it's less about payments transaction fees. It's more about maintaining the role of commercial banking deposits. Does that make sense? Sorry, it's not a very simple answer.

Let's go to Chris next before moving over to.

It's Chris Cant from Autonomous. I have one on sort of the longer view look on the landscape, I guess, and then one about risks. So Charlie, just following up on something you just said and thinking about what you were saying during the presentation around the number of software engineers and how much that's sort of becoming an intrinsic part of the group staff base. Where do you see the U.K. banking landscape going over the next 10 years? I mean, are we just going to see essentially all of the smaller players drop away as they're simply not going to have the scale of customer base to make things work in the way that you're expecting to. Do you see a small -- do you see challenger banks in any meaningful sense existing in kind of 10 years' time? Or will it actually tend to what we see in software markets, for instance, where it really concentrates ever more in the hands of a few large players. So that will be the first question. And then second, in terms of the opportunities outweighing the risks, obviously, fraud risk is quite a problem, I guess, around AI. And particularly in the U.K., the banks are on the hook for APP fraud in a way that simply isn't the case elsewhere. So how should we be thinking about that? There was a statistic in there that you've saved, I think, GBP 200 million or something in one of the footnotes over the last 3 years I can't remember exactly what I said, but there was a number in there. But how much worse is this getting? And are you worried about it being something that could outpace your ability to kind of protect customers because you can only put out so many tick-box screens, I guess, in the app saying, yes, I'm really sure I definitely want to make this payment. How are you thinking about that risk?

Thanks, Chris. I think the strategic landscape, probably both Ron and Charlie, have probably got inputs onto that one. Should we address that one first.

Yes. So I'll go Ron, and then I'll hand over to you. Look, I think actually, my IR team will say we're going to straight here into our 2026 plus strategy, and I need to be careful. So -- but it's a great question. And of course, whatever either of us say now almost certainly be wrong. I think that's what you need to start with I've been -- as you know, I've been in every major tech disruption geography in the last 30 years, including Silicon Valley '99, 2000 and Asia and various other countries. And what actually ends up happening is almost certainly different. But I will come back to your question. However, in every situation, I've been in the '90s, 2000s and 2010s and hilariously in London, New York and the '90s around trading floors in Silicon Valley during the dotcom boom and then in Asia when Asia really took off in China and Singapore, it was very easy to work out what the basis of competition was going to be. And so I think we have a very clear view around what we need to do to be successful, how the industry shakes out exactly, I think, is less clear, and what its worth without going back to it, using -- having data and data at scale around the area you want to provide advice in the context of agentic AI with a very, very strong digital engagement platform that the regulators understand and trust. I won't go to all of them again, Chris, but those are the foundations that I think we are very clear you need. How do I think the world will turn out in 10 years' time? I think your question is a really good one. And I think financial services may be a bit different from other technologies, partly because the regulator plays a role in this, and they require a level of openness, not just through open banking, and also which will allow probably more competition than you might see in the tech sectors. But similarly, we'll put constraints on the pace and level of innovation because of regulatory obligations, which will be complex for us and other participants in this space. But I do believe people with the best data and the best talent and teams to be able to use it will have an advantage in that future. I alluded to it, but I think there's an opportunity for players to reposition themselves directly with customers. Again, so the intermediation market and model will evolve significantly during that period, which is an opportunity, I think, for some players. And then to go one step further, which I'm sure you already at, Chris, I know you and your organization are incredibly thoughtful about this. We're already -- we need to get ready for a world where apps and what we think of as the big tech today may no longer be relevant. And when you look at the way agentic AI, or more importantly, generative AI, but the ChatGPTs and the Perplexities are emerging, they're going to be a real challenge in this next period for the Apple and Google platforms. I think Google is obviously doing a lot to respond to it. So the exciting thing there is that's going to change this dynamic at the front of our engagement with customers, which has been emerging in a pretty duopolistic type way in the U.K., and it's going to provide opportunities for us to be more relevant, but it's going to be more complex. So I don't -- I won't give you an exact answer. What our view is, is there are very clear competencies and capabilities that will enable us to compete in that world. And I'll tie it back to one thing in this phase of the strategy, if that's right, so to become a bit more real -- in 2022, we said to you on our retail businesses, and it's true on both SME and also insurance, but let's just stay with retail because it's easy. We need to be a leader in both having relationships direct to our customers and building out the best relationship bank, which Ron has talked a lot about and being able to bring the full breadth of the relationship with those customers that want to have a full relationship with us. We know in the U.K., most people have multiple banks, but at least the majority of the relationship. But we're also #1 in every product, which is intermediated. And so we're #1 in mortgages, #1 in cards, #1 in loans, #1 in transport, and they are highly intermediated markets. And so we are also today the best at integrating our services and embedding them through brokers, sharing our data. If it's our customer that touches us through a broker, we can give them preferential pricing in a number of these channels, we're pretty sophisticated out, right? But it's worth, it's the most sophisticated market in the world, maybe buying Ali and Tencent in China, but even there you did the same stuff as we're doing here. So in that world today, we said we're going to be the best at both of those things. We're going to have to do that in the future. And the exciting thing is there's almost a new channel, if you like, which is agents talking to agents and without getting too conceptual and that's how we're going to win. Well, if I just take all of the content away from you.

Now maybe just one build, right? Because indeed, it's a very important question, in terms of can all these relatively smaller players follow. Can they lead? Does it scale? The level of investment needed to build a beautiful ecosystem of agents talking to agents that Charlie was alluding to. And this brings me to my one build, and that is there is only so much talent in the market, even if you make use of vast talent pools like India, Philippines, et cetera. There is only so much talent, and we're all competing for that, and people make a conscious choice for whom they want to work. And I think we can see from both graduates as well as experienced people in AI. People want to work for companies that have a clear purpose and that are winning in the market. And that won't make it easier for smaller competitors in the market to survive. That's my only built.

Ron, do you want me to do the fraud one?

Yes.

Look, this is hugely important. Chris, as you know, actually, the APP fraud is primarily linked to faster payments. It's one of the things that I think is slightly challenging in terms of the way the regulation works. Consumers and businesses get different outcomes depending on what their payment is, credit cards have chargeback and retrievals and fraud built within APP fraud applies to faster payments if you do CHAPS and banks, just use the language, you don't get any fraud protection. So it is one of the things I think we'll have to evolve over the next period of time. And critically, you pointed to the work we've done, and we're really proud of this for our customers, first and foremost. And then obviously, for our shareholders, we have done a lot to invest in AI tools to manage forward better. And it does create friction for some customers. I feel like I should apologize if I'm in front of the media for that, but we know it's actually saved and materially protected customers from being exposed to fraud and reduce the cost to shareholders and we track our share of interactions and transactions, volume and value relative to our losses on frauds, and we're better than the market. So it has been a very significant part of what we do. We had a very big deployment what it's worth in the last 6 months this year where we applied all of those very advanced fraud tools also to debit cards. And so we're protecting more customers with more pace and saving money for the group. Your point, though, which is how does this evolve? And how do we continue to do it? It's top of mind. We do it by continuing to invest and be ahead of how we protect customers. And then I'll do one more link, if that's all right. Ron, it's why we're excited about GBTD, the very first use case that we piloted already, and we proved works between a number of banks in the U.K. was for online marketplaces and actually I know you've heard me say this before about 40% of all crime in the U.K., individual crime is financial crime and about 80% of that has been facilitated by the big tech companies and the biggest facilitation is online marketplaces. It's actually the Facebook marketplace. So one of the first use cases of the programmable money that we'd like to go live with on a system-wide basis means that you could pay money to someone online on an online marketplace, the money would leave your account. It basically goes into escrow, but you -- and so it's available to the seller of the good, but you only release it once the goods have been received. And the reason I give you that as an example is suddenly, if you have programmable money, which means nothing to anyone until you see the use cases, you can see that we can build use cases into this money that makes it safer, and it's actually a tool for managing fraud. Now that won't solve all types of fraud as you know, better than me, Chris. But it's -- there's a huge opportunity to use this as a platform to innovate and our intent would be to make this a platform that's available to the whole market. So everyone can start innovating off this. It will be seamless and create a singleness of money if we're here with the central bankers for the whole of sterling. And I think it's a really important tool for us to then be able to innovate and for the whole market to innovate on how we continue to fight fraud. It's not a simple answer -- single answer by itself, Chris, but it's another reason why we're excited about GBTD. Thanks for the question.

We'll take a question from Aman first in the front, and then we'll go on.

Amandeep from Barclays. Look, I appreciate you're not going to want to talk about your new strategy and your new plans and forward look. But clearly, this technology has potentially compelling implications for the operating model and the number of people that work at an organization like Lloyds profile of that workforce and ultimately, the cost base. So is there anything you can kind of give us, perhaps not in terms of numbers, but qualitatively, how you're thinking about the evolution of these from here. And the related question is just around the role of the branch in this world, particularly if you're opening customer current accounts in minutes, interested to kind of get your updated thoughts on what the value proposition of the bank branch is now?

Thanks, Aman. I suspect it might be quite limited as to how much Charlie can respond on these questions, but I'll let him respond on both of those areas. .

I think it was guidance from [indiscernible] guidance to me about the guidance I gave. Look, I think the first thing to say on your point about operating leverage, whether it's cost efficiency and the mix and type of people is it's actually been a meaningful part, not generative AI because that's still newer, and Ron gave you the numbers. It's been a meaningful part of this phase. The biggest thing is where you can enable customers to serve themselves, whether their businesses, pensions customers, retail customers and everything in between. And then once you've got that, if you can help them get whatever answer or question they've got right first time in a channel of their preference and increasingly, 97% of our interactions are digital. The U.K. is by far the most out of Western democracies by far the most digital country in the world that way, even compared to Nordics and compared to Australia. It's quite advanced, obviously, U.S. and Canada behind. So that's been a big part. And as you saw in the presentation, we're saying that 70% of the GBP 1.5 billion of additional revenues by 2026 are being supported by this and 60% of the GBP 1.5 billion gross cost saves to date, we haven't given you a target for 2026. Our original target was GBP 1.2 billion is what we committed to by 2024, have been delivered by those kinds of efficiencies even before we think about the next stage. So that's the first thing. And so you should absolutely expect that kind of scale of impact. The exciting thing is when you think about the new technologies Ron's talked about on one of the slides, we talked about one of our Agentic AI use cases, which is supporting our contact center and customer service colleagues, that's been live for a while now. Ron, I think it was Q1 this year, end of last year, we went live with Athena.

Q1.

Q1. We've got tens of thousands of colleagues using it, and we can track the minute saved per interaction per colleague. And as you know, in those environments, you can then manage and optimize what you need to support customers or you can take it as a trade and say we're going to materially increase how we serve customers and provide a higher service level. And those are choices that are operating at a scale that we're using today and will be increasingly capable and available to us. So we definitely see it as a really significant opportunity to differentiate our propositions like in the investment propositions we talked about, drive efficiency right across the organization and manage risk better in the spirit that Chris has laid out. You will forgive me if I won't give you that guidance at this stage because I don't want to go beyond our current targets of 2026, and I want to deliver those. But the confidence you should have is these are the kind of mechanisms we've been using to deliver that growth and efficiency to date. And we see these new tools now that we've got the technology and the platforms built as giving us the ability to kind of continue that journey and accelerate it in a number of areas. What I am most excited by for this work is obviously, I care massively about the operating leverage and the efficiency. But our ability to differentiate what customers experience. There's even simple things, right? So why should there be any wait times in IVRs any longer when you've got brilliant colleagues with the expertise that can support around the most difficult transactions and you've got the agentic AI capability to support around the simple stuff. Can you start to eliminate this isn't a commitment by the way, Douglas, don't get worried. It's not a commitment around what we're going to do, but there's even simple things you can materially improve what customers experience. And then when you get into providing truly what we have wanted to do and certainly I've been and committed to for over 30 years, getting down to the customer of one and actually differentiating how we support those customers. We've finally got the technologies to do what I spent 35 years doing. Ron has been doing it for even longer, but he was in the wrong industry to start with. So I'm sorry, I'm not going to answer that question fully. Just in terms of the role of the branch. Look, I think, again, you should just see this as a natural extension of what we're doing. The branch plays many roles. And part of it is about community engagement and us being relevant to our communities. And it's one of the reasons we're differentiated. We haven't talked about this, but -- there's a lot of that noise in the market place about market shares, but we have grown our market shares of personal current accounts and business current accounts in this phase of the strategy. And that's where the value is and that's where the trust is at the moment that we're seeing, obviously, neobanks and fintechs take quite a big share of personal current account flow. And I know there's lots of concerns around, can you compete with what's happening? What we focus on is real customers and value, and we've grown our market share for the first time since financial crisis actually. So branches play a role in that. They play an important role around supporting parts of society that aren't yet digitally active. And we announced today a really kind of exciting but also quite challenging. I don't know if maybe you had a chance to look at its survey around digital engagement and the pace of adoption of generative AI and digital has exceeded even my expectations in the last 12 months, but there are still 5% to 10% of customers who don't engage digitally. And for what it's worth is when you look at where the value of -- in the U.K. retail market is and in the commercial market, SME, a lot of it still resides with people who aren't that digitally engaged. So that's important to recognize. And then in the moments that matter, people still do want to go talk to people. So I think you've seen a significant, really significant reshaping of our branch network, ignore points of presence for a second, the stuff that we are most excited about is that we managed to virtualize, if you like, the RMs and the relationship bankers so that they're not stuck with a single branch and they're available to talk to customers and bring the expertise they have in a much more joined up way. And I think that's going to be the really exciting part of the next few years, which is how do you let customers engage with our intelligent agentic AI agents to support them when they are comfortable and that's the best outcome for getting their tasks done or doing what they need to do. And then when they need to talk to somebody, how do we give them access. And in some cases, I still think a physical presence, which will be very important in this period. So we have got a kind of going back 10 or 15 years, everyone used to talk about multichannel branch banking. I've banned that phrase since I've been here because I don't think it's very helpful. But if you take it truly customer back, that's what's driving our branch strategy. We still believe it's important. It's very important for some customer segments that are valuable for our shareholders. It's a very important moment. It can be a way of providing support in moments that matter. But clearly, we need to significantly modernize and drive the operating leverage out of that infrastructure, and you've seen that happening in this space.

That was actually a much more fulsome answer than I expected.

Or wanted.

[indiscernible]

I know we're at an incredibly early stage, but and it's really connected with the previous questions around challenger banks and branch network. But I was just wondering, and it's kind of a super high level, but just how much more successful the selling machine do you think the app could be then the -- compared to the branch network and the old way of doing things with adverts. And specifically, I was wondering if you had any data, it would be really interesting to see how much longer people are spending on the app now with all the new offerings that you -- that have been coming on in the last few months. And maybe you have some data about how many -- what proportion of customers are using agents and perhaps how that compares to the time in the branch? And related to that, it would be interesting again to see just how successful -- what the success rates are like on some of the personalized offerings? And if you've got data, you can compare that with to the old letters of the branch network. And I guess, wrapping it all up, just into kind of longer-term aspirations. I mean what do you hope -- how do you hope to increase the cross-sell or the kind of fees you hope to generate per customer as the app develops?

Charlie, meeting customer needs, do you want to start that question and then we'll probably layer on top?

Yes, Ron, you -- so the first thing on this, I think the first question is, how much better is the app at selling than branch? And how does that evolve? The answer is even today and Ron, you share the data, the majority of our I would like to call it needs fulfilled rather than sales. But a majority of what we do is through the app and through digital services and then through our very important intermediaries, where they remain a really critical part of serving our customers. And the branch is true actually of most financial institutions in the U.K. The branch has become less focused on sales since the financial crisis, right? And that's the strength, I think, because it gives us much more scalability, much more reach and then we can be very innovative around how we position ourselves, whether it's bundling through social media, through the Internet and pulling people into our journeys or into the app, and all of the above. So the majority is already happening there. And the second thing is when we do personalization, and we haven't shared this data. But I think what I might just do with this question is we might take it into when we come back on the next strategy, think about how we could share some of this data. So I'm not going to give you the specifics, although I know the actual data. You tend to get -- I'll give you the anecdote 10 to 100x more likelihood that someone will engage on an offer or a discussion and go into a meaningful journey if it's properly personalized and I use those 2 phrases contextual and relevant and relevant and contextual means it's presented to them at the time they're considering they need a product in a channel that they want to engage with you on. And by the way, even, and I always laugh I use myself, I like to -- I travel on a train at 6:30 in the morning. If someone sends me something as an SMS at 6:30, I might look at it if you send it to me in the rest of the day. I'll try not do it in front of you, but I won't have time. So we can start to personalize at which channel, what time of day, what tone of voice and then based on something you've done. You're searching the Internet, you've looked at our channels. So that's contextual and relevant you get massively, massively higher engagement rates and then you have what's called a funnel fulfillment opportunity. So these are huge things. We are the leader in the U.K. around this relative to high street banks. The fintechs do this very, very well, by the way, as you know. And we've done a -- it's not for today, but another part of the journey we've been on is to go from basically very limited use and engagement of social media to being one of the leading banks through the transformation of our marketing and data functions. And if it's okay, just because I'm not going into guidance beyond 2026, I'm not going to give you the guidance on how far we would like to go in terms of either cross engagement, breadth of depth of relationship or revenue growth. However, you know we laid out at the start of this phase, and we'll give you an update next year on depth of relationship as a good example of a metric we said was important. We gave you metrics at the start of this phase, and we're on track to deliver them around numbers of FTEs per customer from a distribution perspective to give you this feeling of efficiency growth that we're talking about. We gave you some step changes in 2022. We'll continue to report on them, what I can tell you now is we're going to meet or beat those targets. And I think those things should continue to be value levers, horrible consulting world that we will prioritize and focus on going forward. I'm sorry, I can't answer it more specifically, Max, I just don't want to get to guidance, but it's exactly the right question. I love it. Let's come back and make sure we give you more clarity going forward.

Ben, in the front line.

Ben Caven-Roberts from Goldman Sachs. Just 2 questions, please. First, on competition. And then second, a bit of a follow-up on costs more conceptually. So within competition, it sounds as though the greatest part of your competitive moat revolves around quality of relationships, depth of data. So first of all, would you agree with that characterization? And then secondly, if you think around potential for disintermediation or potential areas of your competitive moats that you're trying to reinforce the most to ward off competition. What would you think about there over the medium term? And then secondly, just in terms of costs, conscious you don't want to give any numbers, and that's completely fair, but more conceptually around the shape of tech investment. Is that changing at all the type of products that you're investing in, the type of tech change that you're trying to achieve? Is that changing in the percentage of OpEx that tech is? Is that shifting as well even if the gross saves might be increasing?

Thank you, Ben.

I'll take the first, Ron. Do you want to do the second one.

Yes.

Is that -- because they're both great questions. Obviously, as always, Ben. Look, on your first one, I think the answer is yes, but I'd say that. I'd need to think about it more. So we definitely think quality relationships and depth of data that underpins those relationships is important. To build on quality because quality is -- it's not just do they trust that? Do they have a relationship with us today? And do we have a channel. I think trust, we do think trust is today, fundamental to who we are, that may sound even less quantitative and quality. But the reason I think it's important is when we look at where do we have the deepest relationships with people that bring the most value to us and that want to then work with us going forward, trust is at the heart of what they say they care about. And interestingly, as you start to move into an agentic AI world, trust is more complex. And people are already highly aware, again, our digital survey, we've just announced is really, really interesting in this context. They are willing to engage on very emotional personal issues, but they are highly aware that they may not trust the outcomes. And as we start to talk about regulated financial services outcomes, we think trust really is an important part of this. Now we spend a lot of time torturing ourselves on what trust looks like, how you build trust and then how you manage it, especially in this world that Chris talked about, where fraud is prevalent and where historically, banks have seen that they're not always trying to help you get the best outcome. We spend a lot of out of time helping customers get the best outcome on our basic services today and protect them from fraud to build that trust. And we do think that's going to be a really important battleground in this next period of time. And Ron talked about a whole bunch of other stuff like our role around ID&V, identity and verification as one of the mechanisms for being a trusted provider of that data. It's particularly important, we will talk about, so I'm sure in the strategy, but we'll have to see how this evolves if you start thinking about agents giving financial advice. So I think I agree with your question, but the build would be trust, I think, is a really important part of it, and we need to be the best at doing that. And of course, by the way, with breadth and the scale of your starting point, both informs the data, but also your ability to then build and test and pilot things and then iterate at pace. So I think that would be the third one. And then potential for disintermediation, look, this has been the history of our whole working careers, right, which is you've seen fragmentation in the value chain across wholesale retail payments and insurance, and then domestic and international payments. And you've seen, especially as we've talked about in this market, the emergence of a significant amount of intermediaries on some of the most important products. So I think there is already a significant amount of disintermediation, and you have to be the best at working with the people that are intermediating you to still succeed and be relevant and to work out where you add value versus them. And I think that will continue. What I teed up earlier is I do think this next phase, and we'll think about it strategically provides us an opportunity to get closer to our customers in a number of segments and product areas, and also to join up for customers in a way that makes it much, much simpler for them. So that's one of the strategic objectives we will have and we do have today. And without being too theoretical, the Home Hub is just a brilliant example of that. And the question I got earlier around what percentage of the Home Hub customers doing their refinancing with us. Remember, 85% of mortgages today for the market are going through brokers. We actually outperform in our direct channel were up about 24% of our mortgages that distributed ourselves. That's always on the new flow, the remortgage journey and engaging customers and making it really simple and showing them what is value if they remortgage with us, that happens digitally is a great way of broadening and capturing the value from an existing relationship is intermediated. And so that's been part of our strategy today. Right. I'm going to hand to you, Ron, for the tech and cost discussion, which is a great question as well.

Yes. I think this is spot on, right? And as you heard me speak about the first phase of our transformation, much of the investments went next to, obviously, improving our customer propositions, you saw the examples but a substantial part went into building new foundations literally brand-new data center, our digital platform, our public cloud foundations, our cybersecurity platform, our data platform, all of it, right? So quite foundational, which I would say has a longer-term return on investments across the group. I think looking forward, and I'm not seeing -- we're not investing in those foundations anymore, but the shape will change. Obviously, investing more in the AI space and investing more into individual propositions, addressing more specific customer needs to Charlie's point, in the end, you could imagine -- in the end, you could imagine that while using GenAI and agentic AI, you would be able to specify and customize your offerings to the segment of one, right? So yes, the shape of our investments and the kind of type of investments will change over time. And the second element to this is -- and there was a number in the slide saying gross savings on delivery, and that number will increase as we further improve our productivity in the tech and data domain as we improve the skills of our workforce while we implement agentic AI adding further to the productivity, you get more value for the money invested. So even if we would leave the investments at the current level. And again, that's for Charlie to discuss next year, even if you would kind of maintain the same spend, you will get more value for the money that you invest.

Okay. Thank you. The seminar is proving very popular in the fact, I have numerous questions online as well as those in the audience. A number of the questions online have actually been addressed through other questions. However, particularly given your discussion, Charlie, on trust, I think there's one question here that is probably worth addressing at this point. It's probably one that you're quite close to as well. The question was, how does the executive and the Board ensure ongoing accountability and oversight of AI systems particularly in high-impact areas like credit decisions, fraud detection and customer profiling?

Yes. Maybe I'll say one thing and then Ron, I know we shouldn't double tag questions, but you've built out the team around ethics and oversight and AI, and there's some really good developments in that context, which is like the controls and the capability to do this. But it's a really important question. Look, the starting point is that you have a Board and executive and then below the executive, the people that actually understand technology, data and AI. And as you know, we've done a lot in the group to refresh the talent, Ron is an example. You've met some of the rest of the GEC, my group executive committee and a lot of them have a lot more depth around technology than you'll see in other organizations. And that was a deliberate choice because we knew to deliver the strategy for our customers and for our shareholders, we needed to have that capability. The second thing I think is you need to invest in the experts who can then oversight and manage the different tools we're talking about, and I'll ask Ron to talk about that in a second and have the PhDs and have the systems and the tools that enable you then to do quality checking both pre-putting AI tools into production, but then critically for the ongoing optimization and review and control of them. I'll talk about that as my third thing. And as I said, we started -- we start this next phase of AI with already 800 live AI models using large language models and other form of AI infrastructure. So this isn't our first rodeo. We know how to do this. We know how to do it safely. And to give you a feeling for it, one of the things because all AI models have a level of black box outcome is you have to get very good of back testing and the complex part of that is creating a clean data set against which you can back test. So we have significant capability around control testing. The third thing, which I won't spend more time on now, but it gets -- it floats my boat because I spent my old career trying to do this. But I just teed it up earlier on investments. We've got very good as an industry of dealing with humans, which are black boxes and doing 1% or 2% quality checking retrospectively on their outcomes. What we're already doing with Ron's team's input is creating agents, monitoring agents who then audit themselves produce exception logs, you can have agents looking at those outcomes with 100% fact checking and then presenting data for a human to then oversight. The world that we have built, the biggest retail and commercial bank in the U.K., where you do 1% or 2% checking with a lot of humans that are black boxes. I love my colleagues and they're awesome to one where you can start to have more data-driven oversight is the one that we are able to start executing today and it will become the model that defines great quality control, whether it's credit decisions or whether it's processes or whether it's customer outcomes. So I think that's the third thing. Ron, you want to just want to talk about the capabilities that you've been helping us build out?

There is actually 3 layers to this. One is people. The second part is capabilities or technology capabilities. The third one is governance. So on people, we have a specific data ethics and AI ethics center of expertise, like Charlie said, very smart, well-educated and conscious people. Secondly, we, like I slightly explained in the presentation, right? AI use cases require access to data, access to LLMs, and they require guardrails, which translate into real capabilities to avoid data privacy breaches, intellectual property breaches, regulatory compliance breaches, biases, hallucination, unethical behavior, all of these guardrails you need to build into your platform, which is arguably the most challenging part, giving access to LLMs or unlocking our own data, I would argue, are the more simple part. So that's on capabilities. And as we are building out our agentic platform, these are embedded, crucial and critical elements into that. And thirdly, like I said, to governance, we do have a data ethics committee which is actually chaired by myself with obviously our legal and risk people, business representatives. Where we do discuss the use cases, we want to bring to production discuss what the ethical dilemmas, potential ethical dilemmas are in there and how we deal with them. And it's not the case that every single proposed use case will survive the scrutiny of that committee. We do actually reject certain use cases that we believe are not sufficiently well thought through in terms of how they deal with ethical dilemmas.

And then the outcomes of all those controls will go back up through governance to both our committee at the top of the banking and the executive and up to the Board, which I think was the question...

Excellent. We've already run out of time, but let's just take one final question. We'll take it from Andrew.

I might find sneaking to, if I can, sorry, about that typical. The first one was just you highlighted the ranking in Evident AI and how you've moved up 12 places to 15th. When I look at that index, I mean I think it's very difficult to rank the banks, they do a very credible job of attempting to do so. But the top 10 is still dominated by the U.S. banks. So is there anything for you to learn from the U.S. banks? Or do you simply think that they are better at advertising what they are doing already? And the second one, just on costs gross versus net. I think you talked about agentic AI being almost a continuation of what you're already seeing evolving in terms of digital adoption. I mean if you were to go about 15 years ago and tell somebody that the branch network across the U.K., not your industry would halve, I think they'd be surprised. If you were to go back and tell me halve, but costs will be broadly stable, they'd probably be even more surprised. So what do you think about gross versus net costs from here because clearly, the gross cost savings, which it is sizable, but does it just all get redeployed.

So I do the second one. Ron, do you want to do the first one?

Yes. Sorry, Douglas. Sorry, remind me the first question was on...

Evident AI, U.S. banks and lessons to be learned from that.

If you -- and we did study, obviously, the report in depth and then since we participate, we get more detailed insight, a bit behind the curtains, I would argue. Obviously, the U.S. banks are leading. However, if you look at the ranking, the kind of top 3 banks in that ranking and from top of mind, that's JPMorgan, Kepler and RBC, they're clearly leading. They're ahead of us, and we should acknowledge that. From, let's say, the top 3 down to probably roughly where we are, it is more or less we're all on par level. Some may excel in some dimensions, other excel in other dimensions. I think what we strongly believe is while we continue our efforts, and this is not like we explained a small effort of a small group of people. We are investing in our senior leadership together with Cambridge. We are investing in our data science and engineering community to retrain them using agentic AI. And on the broader colleague base, we are skilling them as well, reskilling them with AI tools so they are better positioned for the future. And approaching this in a more sustainable, broader, and you could argue a democratic way, I believe we are well set up to gain more position. It's not per se spending a lot of money, which arguably the U.S. banks have plenty of. Do you want to take the second?

Great. Yes. No, thank you. So the second one, we've had this discussion with this group before about gross versus net cost. So it is the kind of question. Obviously, the first thing I'm going to say is one you'd expect me to say, which is we're going to get to our 50% just under, if William were here, just under 50% cost-to-income ratio next year. And obviously, given the market we operate within, remember, you got to remember because cost-to-income ratios are revenue and cost there's many markets where revenue is -- the margins are wider and what you can do as a bank are very different. So given the scale we've got the market we operate within and then the efficiency level we're delivering, achieving that, we think, is an important stepping stone. So we're committed to that. And whatever your numbers are, that will be having achieved approximately 30% revenue growth since 2021 and having dealt with a very significant margin compression for our mortgage book, which means that revenue growth is showing actually very significant alpha underlying market share growth and underlying business growth and the easiest one for me to describe to you today in that context is the other operating income growth of 8% to 10% CAGR, 9% CAGR now through the cycle, which is very differentiating. So the reason I go there is that's the first thing. Now as you point out rightly, that has shown growth cost increases. And even though we've saved GBP 1.5 billion -- sorry, net cost increases, even though we've seen GBP 1.5 billion of gross cost savings, we've seen 2 things broadly inflation, which everyone can understand and model number one. Number two, I asked for permission from all of you to invest GBP 4 billion more beyond our run rate to do the things we've talked about today and to tee up a business that was pivoting back to growth and transforming and being competitively relevant. So that's been a higher impact on our costs. And then the third thing, which is I always think of as good cost is business growth. And if -- I don't know whether we're going to be able to do this next year, actually. But certainly, when I look at it, I look at the marginal cost income ratio for the business growth I'm getting ex those 2 other factors. And obviously, that's a very, very healthy number given how we operate. What I think is going to happen going forward, I'm not going to give you guidance, but I do think whatever you look at, you should definitely -- we should definitely as an industry, and we will have very high expectations on our ability to continue to deliver gross cost savings. I can't control inflation, but you'll have your own models or views around that. Ron talked to you about. I'm not going to give you guidance on it. But our ability to deliver more with the same capacity or less, we think is strategically really important. Asia, we can compete from a pace perspective. But secondly, so we can deliver more for less to our shareholders and investors from our technology capacity. So that will be part of the story that we have going forward. And then the third part, which we need to come back and be really clear on is there is some, I think, really exciting growth opportunities for this group. We are proving we can now capture market share, be relevant and create value. And so -- and those will require investment in people and capabilities to grow and we should be transparent around how much is actually growing the operating expenses in support of accretive growth. So those things will be true. What happens on a net basis, I think will depend obviously on the combination of choices we make and for the industry more broadly, the choices they make for us. But as you know, through '26, and we'll come back beyond that, this is a business that you'd want us to be able to invest in because the returns are very, very strong. Sorry, I can't give you the full answer. We will take it as input for next year.

So thank you. So given the time, we'll need to finish the questions at that point. Before we actually finish, if I could just hand over to Charlie to put some brief closing remarks.

Well, thank you, Douglas, and thank you, Ron, for joining. Look, I really appreciate you joining the session. I really appreciate you doing it in person. I hope it's been interesting -- as I said at the start, this is the last of the kind of seminars alongside our quarterly updates for this strategic cycle. Thank you for bearing with us. We knew this one was a bit more conceptual, and the risk was Ron and I are deeply, deeply passionate about technology innovation and we've spent our careers doing it. So thank you for bearing with us with the more theoretical answers, but I hope it was helpful for you. Obviously, we're working towards the strategy update next year. I know we haven't given you a specific date, but we have said that will be in the summer next year. We'll come back and give you that date probably at the year-end, I'm guessing, which is at the end of January because we're doing prelims to get pace going and get the organization looking forward. So thank you very, very much for attending today and really look forward to the next meeting, I suppose, which will be the prelims in January. Thank you.