EY helps clients create long-term value for all stakeholders. Enabled by data and technology, our services and solutions provide trust through assurance and help clients transform, grow and operate.
At EY, our purpose is building a better working world. The insights and services we provide help to create long-term value for clients, people and society, and to build trust in the capital markets.
How GenAI is transforming operations in APAC banks
In this NextWave Banking in Asia-Pacific podcast episode, speakers tackle how banks can leverage GenAI to enhance customer service, streamline operations and stay ahead of fraud and scams.
Join the conversation as industry leaders Lukas Bower, EY Asia-Pacific Financial Services Generative AI Leader, and Andrew McMullan, Chief Data and Analytics Officer at Commonwealth Bank of Australia (CBA), delve into the transformative impact of generative AI (GenAI) on the banking sector.
The discussion centers on how banks can leverage GenAI to enhance customer service, streamline operations and stay ahead of fraud and scams. The speakers also touch on the ethical use of artificial intelligence (AI), workforce upskilling and the importance of a customer-obsessed, AI-driven organizational culture.
Key takeaways:
The AI journey of CBA and the challenges they faced
The strategies that CBA implemented to create a seamless digital experience for customers
For your convenience, full text transcript of this podcast is also available.
Andrew Gilder
Welcome to the next season of the EY NextWave Banking in Asia-Pacific podcast! I am your host Andrew Gilder, EY Asia-Pacific Banking and Capital Markets Leader. GenAI is among the powerful technology-enabled forces which are shaping the banking landscape today. Taking advantage of the transformational power of GenAI requires a combination of new thinking about a longstanding challenge for banks — how to innovate while keeping the lights on. But banks clearly understand the urgency; a huge majority are already dedicating resources to GenAI. For today’s episode, we swing to the EY office in Sydney where 2 industry thought leaders are ready to get into conversation on how the value of GenAI extends beyond just process improvements and cost savings to actually strengthen their capabilities and foster more and faster innovation.
Lukas Bower
Thanks, Andrew, and hello to all of our listeners across Asia-Pacific and the world. My name is Lukas Bower, and I'm the EY Asia-Pacific Financial Services Generative AI Leader. Today, I am privileged to be talking with the great Andrew McMullan, who is the Chief Data and Analytics Officer at Commonwealth Bank of Australia, which we also know as CBA. Andrew, welcome to the podcast. How are you today?
Andrew McMullan
I'm very well. Lukas, thank you very much for having me.
Bower
Great to hear. Let's get right into it. To start, maybe tell our audience a little bit about CBA.
McMullan
Yep, no problem. Commonwealth Bank of Australia is the largest financial services provider in Australia. About one in three Australians would call us their main financial institution. We're also leaders in home lending space, household deposit, credit cards, and in business banking, as well. So, the latest data story that I saw, we were the 12th largest financial services company in the world by market cap.
Bower
Andrew, CBA sounds like an amazing institution and you're doing a lot with AI. So, tell us about some of the primary drivers that are motivating you and your team to implement AI?
McMullan
That's a very easy question for me to answer. At Commonwealth Bank, we have an incredible obsession to serve our customers exceptionally every single time and every single channel. So everything that we've been thinking about since the explosion of ChatGPT and generative AI, how do we use this to help our frontline serve our customers better, how do we use this to create more compelling experiences for our customers across our digital channels, as an example, how do we just automate and improve every service that we offer to our customers, so it just feels like it's natural in the moment real time, personal, etc.? So, everything that we've been focused on touches upon that customer obsession.
Bower
Excellent. So that's going to transform your customer relationships, is it? And how do you see that working?
McMullan
There are a few ways that we would see that working, let's kind of split them into the different categories that we think about. So, I mentioned that frontline staff who serve our customers, one of the really interesting programs of work that we started with in generative AI was, and it's a really interesting story, when you think about, customers go to a branch than call the call center. They probably think that it's quite simple for the staff to be able to get access to the information that they need to serve the customer and answer the question that they have. And when we started to look at the repository of information that our frontline staff have, if I think about the retail bank, in particular, we have a digitized repository of all the standard operating procedures, the product information, etc., services that we provide. There were over 4000 documents in there.
McMullan
I stood back. That was one of the learnings that we actually stood back and thought, wow, how did we ever expect our people to be able to get across all of that information and understand everything that customers would need given that every customer comes in with a different query or question. So, the first thing we did was to create the ability for generative AI to help our people get the answers to the questions they had for our customers much more easily. And we're seeing really great results with that. Our frontline is really enjoying engaging with that. It makes the job easier. And making your job easier to serve customers and making customers happy is a really great thing to be able to do. So that's one use case, which is to help our staff be more productive and serve our customers better.
McMullan
And then the other area that we're really focused on is across our digital channels. When you're looking for information, sometimes it's not always easy to get it. And again, thinking about all the products, we're a full-service bank. So, customers will go to the app and they're looking for things. Clearly, you can't have everything that they're looking for available on the front screen, so they have to search. What we've been working on is how do we create a user interface for our customers that when they go to the app, they can find everything that they need.
McMullan
And we have a very ambitious program of work here where we believe that if you think about ChatGPT and when it launched, I think within the first three months was 100 million users, within the first six months, almost 200 million users, and not one user guide on how to use that. So, we're very excited about the ability of our customers to be able to engage with that type of capability and get the answers to everything that they need answered in the app to do that quickly, to do it accurately. And then eventually we see the evolution of generative AI, not only being the return of that information, but the large action models will start to action what the customer needs us to help them with.
Bower
Amazing. That sounds really exciting. Chomping at the bit to hear more. But I know it's not all wine and roses, right? So surely there have been some constraints and barriers along the way, what have been the key challenges you faced thus far?
McMullan
Some of the challenges, I suppose, because it's such a new and evolving space, making sure that you've got access to the right talent. I would say, probably being in Australia, we're a little bit removed from where the action is. And a lot of the big companies who are investing there, so if you think if it's Microsoft, or AWS, or Meta, or Google, a lot of that's based in the US. So, in the early stages, as it was evolving, we find that we had to connect with those companies a lot more than we’ve previously been doing, which has been helpful in the relationships, but also really interesting for us to understand how fast things were moving. So, what we have been really focused on is how do we make sure that we can invest in our own people but pick the right partners to make sure that we stay up with the pace of generative AI as it evolves.
Bower
So it's great to hear that you've overcome some of those key constraints. Andrew, how are you turning your capability into some specific use cases, if you don't mind sharing your secret sauce?
McMullan
Yeah, it’s actually even linked to the last question. Part of the reason that I think the organization's trying to move so fast in this space is, because of some of the success we've already seen, one of probably my personal favorite stories of how we've been using AI and now generative AI, is how we're protecting our customers from fraud and scams. There are two sides to this actually when you think about generative AI. On one hand, the scammers are getting more sophisticated. So, they're able to use generative AI to send more messages that look more like they're coming from the bank, etc., to trick more customers into being frauded or being scammed.
McMullan
So we've been focused incredibly hard for the last couple of years to make sure that we can try and stay ahead of that. And we've actually built a new fraud platform, which runs AI models and returns that count AI scores in under 10 milliseconds to make sure that we're alerting our customers if we've got any suspicion of a transaction being a fraud or a scam, identifying scams really early when they happen and making sure that we protect our customers from those as much as we can. And even in the last six months across fraud and scams, we've been able to help protect our customers from over 200 million in losses.
Bower
That is remarkable. So, looking at that split of use cases, you've got some that are customer facing, but a lot it sounds like you're behind the firewall more in the ops space. What would you say that split is? Where's most of the benefit being realized today?
McMullan
In terms of the very early stages of the launch of the customer facing experience, that one is a big focus for us and we've got a lot of our effort on it. We want it to be the best experience in a banking app and financial services globally. So, we're running pretty hard at that one. But I think the others are where we're having a lot of success already being shown. And most of that is coming through in how we're helping our people do their jobs more productively, and that might be I mentioned, some of the operational processes, frontline. But also on the Copilot coding, we're getting a lot of great feedback from our teams, and how they're engaging with those types of tools.
McMullan
So if you think about the amount of time people spend writing code, now you can kind of get the base source code really quickly using Copilot, make any amendments to that, test it, deploy it, etc. So, we're seeing good efficiency for our engineers there. And also, we've been building out an AI Insights tool. So rather than having a lot of people coding in SQL or whatever the language is that they use to provide answers to the business. We've created the capability for our business teams to type in the question that they have, and then they get the answer in the format they want.
McMullan
So you might be running a mortgage team, and you're like, how many mortgages that we manage to help our customers get in Melbourne yesterday? Either be like, do you want a pie chart, or a table and you can choose what you want. It'll then prompt what the next question might be with some generative AI to say, well, did you want to split that down by postcode in Melbourne or whatever. And then you can just interact with that. No code is written.
Bower
That's obviously a big productivity enhancer, it must have your business really excited. Going back to the customers, you mentioned Open AI as adoption and how quick that was but there's also now, we're kind of past the peak of the hype cycle, and there's a lot of skepticism. How are your customers responding to your AI vision?
McMullan
It's a really great question. We do a lot of research from CommBank but also engage with a lot of external agencies. We partnered with the Australian government. We’re one of four organizations to write the first Australian AI ethical principles. So, we've been very engaged in that side of understanding AI being responsible how you use it and scale it.
McMullan
I think in a recent benchmarking exercise, there’s this company that’s called Evident, that do an AI benchmarking globally. We were ranked sixth in the world in terms of our ability to use and understand AI, but we were ranked number one in our ability to use Responsible AI and we’re trustworthy in the way we use AI. So, we were very pleased with that. And I think the point of all that is that we've been engaging with communities, industry bodies, and our customers for a long time in how we safely use AI.
McMullan
And actually, I mean, I don't know if you know the story, but we've been using AI at CommBank for over a decade. And we have a customer engagement engine, which is an AI powered system that we use to engage with all of our customers. So, after we finish the podcast, I'm going to make you get your phone out and I'll explain to you how every day there are millions and millions of opportunities for us to engage with our customers. And all of that's driven by AI machine learning.
McMullan
So our customer engagement engine, we started building in 2016. We've talked about it, and our external releases and investor reports every year for the last six or seven years. So, we've been communicating to communities and the industry for a long time now about how we use AI safely. In the customer engagement engine, in 2016, we actually started with one laptop in one branch in South Coast Southern Highlands. We now run a system that's got well over a thousand machine learning models running over 157 billion data points in real-time, responding to their customers almost 50 million times a day across all of our channels, to engage with them and understand their needs in that moment.
Bower
We talked about customers, but what about your workforce? We hear a lot in the market about people concerned about the future, their jobs, being superseded by AI, especially in some of those areas you mentioned: operations and financial crime. How is your workforce responding? And how are you managing the change?
McMullan
A lot of the research I've seen, everything predicts that AI is going to create more jobs than it replaces. So we believe that, too. And a big focus for us over the last nine months, there's been an education and upskilling program across the organization. So actually, in the last six months, we've trained over 500 people who were previously more on the analytic side to move into the world of AI. So we want to bring all of our people who want to come with us on the AI journey.
McMullan
But saying that, there's still a huge amount of work to do in all of the rules that we have across the Commonwealth Bank of Australia. We've got to focus on customer service, exceptional experience for every customer engaged with us. We've still got a way to go there but a number one focus in the organization is the Net Promoter Score. To do that, we believe we've got to significantly invest and use AI better than we have done before better than anyone else. And then our customers will love doing more with us.
Bower
Amazing. Thank you. I think this is the question probably our listeners are most interested in. What advice, given your amazing achievements, would you give to other bankers in the region?
McMullan
There's probably a couple of things. Don't try and do too many things at the same time. Think about the areas in the organization that you're really focused, and you see the greatest opportunity and learn most quickly on those. And then you can start to federate and support the organization in multiple areas. Secondly, educate and go from the top down. So, we're very fortunate at CommBank, our CEO has had the vision of the customer engagement engine, which was AI driven early in 2016. He also had the vision a few years ago, when we invested in an AI company from Silicon Valley. I think the challenge that he said as an organization at the time was how do you help us become a customer obsessed AI driven organization. That was four years ago.
McMullan
So our CEO and executive leadership team, a year and a half ago, appointed a new group CIO as well, unbelievably obsessed with the investments that we need to make to use AI across technology and for our customers. So that helps a lot. If you're not having to convince people that this is an area of worthy investment or an area that we need to build talent and if you have to do that, it's a struggle.
Bower
Great. Andrew, thank you so much for your time and your insights. It's really been a valuable session. I'm sure the audience will have a lot more questions. Any guidance to the audience around resources or areas they can look for more information?
McMullan
You've got to get familiar with using the versions of ChatGPT. So, get on whatever web browser you use and just Google the topics. They evolve and every single week, something new and amazing happens. So, wherever you can find the time, go get in your diary, do some searches, look for information. I think there's so much brilliant material and podcasts out there. It's easy to get the learning material that you need that will help you understand what's going on.
Bower
Thank you, Andrew.
Gilder
You've listened to the EY NextWave Banking in Asia-Pacific Podcast. New research from EY-Parthenon reveals how decision-makers at retail and commercial banks around the world view the opportunities and challenges of GenAI, as well as highlighting initial priorities. Download the report via the link found in the description of this podcast. If you liked this episode, please leave a review to help us bring you more insightful and relevant content. And finally, don't forget to subscribe to our podcast on Apple Podcast, Spotify, or wherever you listen. To learn more about EY, our people, and our latest thinking, visit us at ey.com/banking. If you would like to have a further conversation on what you've just heard, or learn more about joining our team at EY, please contact us.