The paradigm shift: how agentic AI is redefining banking operations

Speaker A: Your company's future success demands customer focused, agile, resilient and efficient operations. I'm, um, your host, Daphne Lucktenberg, and you're listening to McKinsey Talks Operations, a podcast where the world's C suite leaders and McKinsey experts cut through the noise and uncover how to create a new operational reality. Today, we're diving into a topic that will represent a true paradigm shift. Agentic AI in banking. Agentic AI is not just another buzzword. It's poised to become a critical differentiator for banks everywhere. Leading institutions are leveraging AI as a platform to redefine their workflows and business models, transforming the way they operate and the way they're delivering value to their customers. However, the story is not the same for all banks. Slow adopters face the very real danger of falling into what we call pilot purgatory, dabbling in narrow use cases without fully realizing the transformative potential of this technology. In today's episode, we'll discuss the opportunities and the challenges that agentic AI presents and how banks can navigate this new landscape to stay ahead of the curve. I'd love to introduce my guest to this episode. David Denison is a senior partner in McKinsey's office in New York and the global co leader of the banking operations practice for the firm. David, welcome.

Speaker B: Hey Daphne, excited to be here.

Speaker A: And Abalash Sridharan is a partner in McKinsey's Mumbai office and leads service operations in Asia. Abalash, so great you can be here.

Speaker C: Daphne, wonderful to talk to you.

Speaker A: Great. So let's launch straight in. Um, so you know, we talked about the intersection of banking and operations and it being such an important place to start thinking about agentic AI. Um, David, let me point this one to you first. Why and where is agentic AI going to be so powerful?

Speaker B: That's a great question. Banks are one of the most heavy users of what we call service operations, right? It's people delivering services to customers. And AI is well positioned to have a significant impact in how the delivery of tasks and services, um, is done for customers. Um, different than traditional AI or machine learning, um, agentic AI enables banks to not only run deterministic workflows, but also run, um, analysis and tasks that are less structured, that are personalized, and that effectively happen only once. Uh, and therefore AI is very well positioned to help banks. On top of that, if you look at operations specifically after the technology and engineering jobs, uh, is prime, um, for AI and agentic AI to be of help. We estimate that about depending on the bank between 50 to 60% of the FTEs are in some shape tied to UM operations. And the potential that AI has in transforming how the work gets done, how the work gets delivered and the services are delivered is tremendous. We have seen a number of institutions based on our um, 2025 global banking report, uh, in the banking space using AI. Uh, still very early to call it a win, but very promising results on the impact, um, around speed, around cost, around quality and ultimately around customer experience.

Speaker A: So thanks David for that, um, intro. So Abhilasha, now over to you. What is making this landscape more complex for banking leaders who are considering AI? Ah, transformation and why is it taking longer than expected? What are you hearing from clients?

Speaker C: That's a great question, Daphne. As is the case with any new technology, creating value from AI wouldn't be a cakewalk and it's going to take its own time. Uh, successful organizations will need to rewire entire domains across operations. Frontline distribution, technology, data science, risk management with AI at the core of IT and supercharging. The impact versus AI LED applications. Just being a hammer looking for a nail, right? And in fact it's interesting you're asking this question because uh, there is a paradox here. The impact of AI Gen AI Agentic AI led application in the sector has been mixed. Nearly 80% of financial institutions that we work with in Asia report using some version of AI LED applications for impact. But a similar proportion globally report no significant impact on their bottom line. And there are a few challenges to consider um, especially for financial institutions. The first is AI initiatives are driven within functional and business silos with unclear linkage to financial value. The second is chasing impact from gen AI alone which has inherent limitations. And it's important to pair it up with a uh, full stack lens where you blend genai with agentic AI, traditional AI automation and digital applications to ensure that you are maximizing value capture across uh, all of the different domains. The third is deploying narrow use cases and point solutions. It is very easy to pick and choose the easiest possible areas for you to go after. Building a chatbot in a customer care, building knowledge management applications for all of your employees, building a fast credit memo writer for a subset of the businesses and then you plateau right after that. That's essentially where you stymie the impact from AI versus driving end to end transformation of domains which are business backed with AI being at the center of it. The fourth is building LLM applications like traditional analytics model which limit the scope to generate content and give decisions. The last one is limited reuse of AI enabling capabilities which result in poor ROI and slower scale. It is important that most financial institutions Amazon eyes their ability to drive cross cutting applications of AI LED capabilities across the bank. The retail bank, the SME bank, the MID corporate institution and institutional bank shouldn't build standalone AI capabilities because the ROI would then be off the roof.

Speaker A: Thanks Abhilash. That is indeed an extremely complex, um, arena of areas of impact and all of these interlocking areas. So David, let me come back to you. What could a successful agentic AI implementation look like for banks and what is actually the impact that is available?

Speaker B: Yeah, no, that's a great question. This reminds uh, me of the time of automation and RPA that we went through about a decade ago. Our view is that depending on the operations journey or process, we should expect, you know, north of 40%, 60%, 70% of capacity creation coming from AI. Uh, but that is only if a few conditions fully happen. The biggest risk that we have is actually what we call the fingers and toes problem, right? Creating models and capabilities that will only help automate a portion of what individual contributors do. Which is basically why um, automation and RPA failed in the past, right? It was very difficult to automate the capabilities of a frontline analyst. So in order to capture that impact and how that impact looks like, um, a few things need to happen. This has to be a priority. Ah, for the top of the house, uh, we have seen a number of our clients in the US where the CEOs are the ones driving the agenda and are holding accountable their teams, uh, together, right? Business, technology, operations, risk to deliver on this. And no single function or single person will be able to drive this, uh, across the enterprise. It has to be a group of leaders committed, with clear targets and accountability. We expect that as part of the impact, the ability for humans to do their work and scale, it is going to increase. So we expect that roughly 20 to 30, 30 agents will be managed by a single person. And so think of it as an individual contributor having a team of 20 to 30 colleagues working together with them, uh, to deliver an outcome. We expect also that a lot of the operations, back office functions, hr, financial planning and analysis, uh, and so on will get an incredible boost out of these capabilities. And we're seeing that actually happen faster today than other areas. The one that our clients have been really interested about, um, at least in the U.S. given the regulatory complexities, has been around risk management in the first line of defense. Things like rcsa, control management, documentation, model review and so on, are areas where there's a lot of data gathering, uh, and simple initial analysis that can help on one hand, free up a lot of time from the frontline to actually deliver to customers and to increase the number of cases and widgets that get reviewed and finally increase the accuracy of the review. So leaders in the organization only spend time identifying the risks and looking at the risks and making decisions to mitigate them rather than spending 80 or 90% of the time collecting data to be able to have those discussions.

Speaker A: Yeah, understood. Um, and so Abhilash, what actions should leaders take and start to think about to get started?

Speaker C: That's uh, another great question, Daphne. Look. To drive AI agentic AI led transformation, banks will need to unlock five enterprise wide blocks that are critical for a successful operations transformation. It is critical for banks to set a bold bank wide AIOPS vision which is linked to their strategy with clear financial and operational outcomes. In terms of creating a competitive distance, it needs to be very specific and linked to businesses. The second is to prioritize high value domains for end to end transformations which are particular to your bank in the context of the region that you operate. For example in Asian markets, in specific Asian markets, the cost of labor could be cheaper. So therefore for you to create full impact, it has to be a comprehensive offering across a combination of all operational efficiency, risk and fraud effectiveness, better customer experience, better employee experience and a couple of other aspects. And finally, this would of course have to be supported by specific milestones which link up to it. So that's in terms of value alignment.

Speaker A: So Abhilash, as we kicked off this program, we talked about a paradigm shift. You've kind of set out the whole lay of the land. It is indeed a paradigm shift that we are talking about here, uh, for the banking industry. Are we seeing the aspiration and the vision already articulated at the senior levels of global banking. Are there bold movers already or is there more that needs to be accelerated? David, let me pitch that to you.

Speaker B: I think it is very much happening, right? If you see the Earnings calls for Q3 for the large US banks, AI has a front and center role, right? I think most CEOs have a view of the impact of AI already. Many banks have a plan and many others are already executing. And there have been news articles about uh, one of the trillionaire banks, uh, and their strategy and how they're driving change management. So this is very much happening. I think the two potential limitations we have is one, that the technology is moving too fast. So the vision is there, but the tools change very quickly. So that is giving a little bit of pause to some of the institutions. Uh, and the other one is perceived lack of clarity on regulatory guidance that will potentially slow down or prevent some of the capabilities to be rolled out in full until everybody's comfortable that they effectively work and that the quality is good, especially the ones that are customer facing.

Speaker A: Abhilash. Is that also what you're seeing?

Speaker C: So if I take an Asia lens, I, uh, kind of see a wide spectrum as far as this is concerned. Uh, on the one extreme end, and I'm quoting this from the interview that the CEO of DBS had with McKinsey a few weeks ago where she called out that AI is eating the world and the bank's future is an AI enabled organization with a human heart. That's at one extreme. And then as I traverse through the colors of the rainbow, we then have other sets of organizations where we see the natural proclivity towards becoming fast followers while focusing on three to four key domains where they can help establish conviction with the board or the senior leadership, such as, let's first ensure that we have conviction and then we really double down when we pretty much see the entire industry moving in that particular direction. And then you of course have at the very back end, people who still are not really sure if this is a technology which is perhaps expected to catch up and scale and want to buy time. And some of the institutions also have inherent challenges around tech talent, tech infrastructure, et cetera. So it is a wide range that we end up seeing.

Speaker A: Thanks, Avalanche. Um, you talked about the important role obviously of the CEO to set the vision. Um, David, I wanted to ask you to come in here to talk a bit more about the collaboration that we must see between the CIO and the COO to bring all of these strands together.

Speaker B: Yeah. Thank you, Daphne. I think it's a very interesting question because it's not only the CIO, it's not only the COO. Right? But there are other CXOs that have to be involved. Right. In our view, the main, um, goal that the CIO has in all of this is to build the right infrastructure and enable the right capabilities for the organization to leverage the stacking of the LLMs, creating ontology layers, uh, and providing access to the different tools and capabilities for the businesses and functions to build upon. The second role and person that is critical in all of this is the Chief Risk Officer and the risk organization. As you know, banks need to go through model risk management. Right. Every model that is used at a bank needs to be documented, uh, reviewed and approved by the model, uh, risk Management Committee. So therefore having a uh, clear operating model with risk where they understand the risks coming from AI and they can move quickly to approve different AI and agentic AI models that are coming through from operations but also from other areas is going to be critical in all of this. Then our view on the role of the COO is basically work with others, uh, by setting the agenda for the operations organization, figuring out which domains need to be transformed first. Where does it make sense to transform the domains? What is the experience that could be or should be delivered through operations? Leveraging the tools that the CIOs have put in place, leveraging the risk guardrails that the risk organizations have put in place, working with HR to make sure that the right capability building and the right people capability exist to work with AI and so on. So it's not a one person task. The leadership of the banks have to work together to be able to deliver AI impact in operations.

Speaker A: Abhilash, what is your perspective on that?

Speaker C: I think David summarized it very um, well and uh, from what we're seeing in Asia, we three components on this one, right? The first one is the combination of the COO and the CTO working as a two in a box is super critical with the COO identifying which are the top 10 processes across the domains that we would have to go after first to drive measurable operational impact. And for example, at one of the Asian banks we recently took the bank in a box and broke the full bank's operations down into about 600 processes and sub processes across a combination of businesses and value chains. And then the COO's team identified what are the top 10 processes which would give them the highest bang for the buck. And the CTO or the CIO and her or his team essentially ensure that they have all the right data platform, the ML pipeline and governance in place in order to support it. Right. The second area where the CTO also expands further is ensuring that uh, for Mobile first customers, uh, you're able to translate a lot of the use cases and the associated impact across all the different channels through which you end up reaching the customer from the time of onboarding to the time of acquisition and eventually at the time of servicing and uh, renewal as far as clients are concerned. Uh, the third part is on the uh, governance risk and scaling which David talked about and which is where I think the partnership with the risk compliance team becomes very critical. And so the three of them, the tripartite combination would have to come in place for this to be a success.

Speaker A: And that really leads me to the last question, which is when you've got that handshake between the senior leaders, what kind of tactics and programs should they be putting in place to bring the whole workforce along with them? David, what have you seen so far that works quite nicely there?

Speaker B: It is not necessarily different than your big transformation change management approach. Derek Waldron from JPMorgan Chase spoke uh, to our uh, colleague Kevin Bueller a few weeks ago and there's an article about this. Um, they have used a number of different levers to make sure that uh, the employees are understanding AI and adopting it. So for example lots of marketing, making sure that everybody is aware of the capabilities that exist and where to find them. A set of, I uh, would call metrics or loose metrics, not tracking individual usage but tracking usage of tools and capabilities that can give an idea of what is more helpful and where to continue developing versus not finally providing um, a lot of training both to senior leaders and mid level leaders to understand how to engage uh, AI, how to encourage their uh, frontline to use AI, but then for the frontline themselves, uh, to understand how AI can help them on their day to day uh, and how to manage AI responsibly which is very important in a banking regulated environment.

Speaker C: I think David nailed that answer. I was recently having uh, dinner with a COO and a CTO of a large Asian bank and they had uh, given a really memorable analogy. They said look, AI is not the pilot replacing the crew. It is the new engine that makes the aircraft go farther and faster with the same team on board. The sooner you can get your employees to realize the truth you're able to really get AI into the bloodstream of the bank. I uh, really thought uh, that captured the essence of how to democratize AI in a bank. Also using this as a plug to call out that David, myself and a few of our colleagues are about to launch a report on AI enabled operations transformation in Asia which is about to come out very soon. So please read that. And it uh, has a lot of details underneath it.

Speaker A: Fantastic. So clearly a lot of shifting um, platforms, um, a lot of things happening. The door is open for banking leaders to see value from magentic AI. Um, but to truly capture value leaders need to move fast and cautiously at the same time is what I'm hearing you say. Um, thank you for joining me today Abalash and David, um, interesting to see we are really in an interesting moment in time. I would love to invite you back maybe in six months maybe in 12 months to kind of talk about how all of these different areas have advanced. Um, thank you for joining me, David.

Speaker B: Thank you very much, Daphne. And Abhilash. It was great to chat with you today.

Speaker A: Thanks. Abhilash as well.

Speaker C: Thank you, Daphne. Thank you, David.

Speaker A: You've been listening to McKinsey Talks operations with me, Daphne Lucktenberg. If you like what you heard, subscribe and stay tuned. Another great episode starts now.