Securing the AI-enabled Digital Future | Michael Miebach | USA House

Well Sarah, first of all, delighted to be here. It's the first time I'm speaking this side of a church, so really fascinating. I'm Catholic, so this is a great question. You know, there's a lot of dialogue. Where are we? Are we early in the cycle? Are we moving on? Is it overhyped? Are we underestimating the end of the cycle? For us as a company, it's important to note that we have been dealing with AI for about two decades. So not generative AI, that's more recent, we all know that, but for two decades. So as a payment network, we assure that the transaction that has to happen happens and the one that should not happen never happens. And how do we do that? We apply machine learning. So we have a lot of data engineering talent and we have a lot of data in the company. So use cases that we see, and that's probably to a higher degree than many other sectors, they are more on the front end of our business as in products and solutions. Yes, we do the internal efficiency stuff like everybody else does, coding assistance and so forth, but it's really on the product side and particularly in cyber security. So that's a really big part of the digital economy overall and it's a great demand and we have been investing and that has been an important part of our AI work. But your real answer is about generative AI and what has happened in the last two, three years. And there the same logic applies because there's some learnings from maybe other businesses there. So we have everybody else, you know, everybody and their mother coming to us and saying I want to sell you this AI tool and that AI tool. And I know we're experimenting with that, but we've been on a journey for quite some time to really invest in our data architecture and that's the real key, that's the real thing. We have one of the largest private sector data warehouses in the world and we could never, you know, in a fast way use that data. So today when we apply generative AI and we source the data from our own data storage and the one that we have invested, which we can do now instantly and it's with the right lineage and permission and all of that, and then insert external data that are threat signals in the cyber security space for example, what a threat actor is doing, where is fraud been detected and so forth, we put it together. That's a great use case of generative AI where we have put out a product that has a 200% lift in fraud detection efficiency. That's real money. That's real scale because we do this globally. And it's not only real money and real scale for us, it's for our customers. So you can start to see real, very tangible success out of using AI. Another area that's interesting, and I'll spare you too much detail around that, but it's the whole area of customer engagement. Many of our customers, which are the largest brands in the world and the largest banks in the world, they're looking to strengthen their customer engagement. How can an outreach to you as Sarah as a customer be not random and an email that you will delete, but how can it be at the right moment, at the right time, in the right channel? That's all personalization technology that takes your preferences, a lot of data together, and then there are AI models that are very specifically trained to produce an engagement with you that will turn into an action that the respective brand is looking for, i.e., you're buying something. So that's also very tangible. We're playing around with a lot of stuff that is future innovation.

I think what I just said is very tangible. Overhyped when it comes to you're going to invest in this AI tool and that AI tool and it's producing real significant cost reduction results very soon. I think that is a little bit overhyped because you have to kind of reimagine your whole way of working instead of putting various tools in between and it doesn't add up to the full picture. And I think there's a lot more coming in the years to come. But it's being talked about like it's the answer to everything at this point. So I think there's a lot more to do on the productivity side.

Yeah, you know, for a business that deals with money, this is a really critical question. So, quick pilots that play around with your hard-earned money and payments and moving money from A to B, that's obviously a problem. At the same time, technology is moving fast and we cannot look for perfection at all moments. So, how do we square this off? Yet again, it's building on what I just said before, a lot of investment in getting your data right. That is the critical thing because then you can go ahead and you can start to play from those data sources, optimize models in pilot, see if it works, evaluate the performance, do that fairly quickly and then you scale. That's exactly what we do. You know, when we scale it means 160 billion transactions a year so all around the world so we have to be very careful but it's that balance. So there are people that deal very specifically with pilots but they can use this infrastructure we've now built over the last three, four years to basically use our data. So that's a way to deal with that. I think the general question is is it an exclusion that a pilot and speed and perfection are mutually exclusive? It's not because I think if you make the ground investment then you can actually square that off very nicely.

Right. So, our lives are all getting more digital. Yeah, we all live this every day. Quite a few of you are on your phones right now and living that digital life. Anyway, the point is with our lives digitizing, the threats are also increasing. That's not a particularly deep insight. But what's really happening is on the cyber security side is the risks are becoming systemic and they cut across physical threats because the water works can be affected. They cut across geopolitical threats when state actors are involved. They're cutting across societal, the misinformation campaigns and all that is all cyber security risk that is out there. When you look at trends, you know, fraudsters have been around for a long time, but now they're organized crime and they're exceptionally efficient and they're monetizing. This is a business in a highly organized fashion. State actors are getting more and more sophisticated. So that's also, I'm not trying to scare people, but that's all reality. This is happening. And AI is not only enabling the good guys in defense but also the offense. So it's an arms race that's going on right now. And I think what's really important is while we are very excited about AI, we also need to see that it comes with a lot of downside from enabling bad activity. At the same time you can turn this argument around and say AI has a lot of promise and we just talked about some of that. And if you get too worried about the downside then you will never get to the promise. So how do you square these two points off? By ensuring that you engineer a layer of trust into this brave new world that we're building from the outside. So every company should think always-on intelligence at cyber security. We got to, and you know, a larger company can do this. They have the resources to do this and there's a lot of best practices to do this. If you're a small business you're the weakest link in the chain and you're also part of the ecosystem. So we have to help a small business and give them ready-to-do packages to ensure that when they launch an online activity, they might have been a physical-only store, that they actually know how to defend. So there's a lot of learnings that are out there and the good thing is we certainly see that in the western world is there is very strong partnership between public and private sector to do that because just the private sector cannot do this. You have to have public sector enforcement capability to go hand in hand. I was on a panel earlier on with the Europol head and she's like all in. Okay. So we got to find those bad actors, you share, we all work together cross-sector which is the other thing that needs to happen. So I'm very excited about cyber security as an opportunity for us as a business. I'm also excited about it as an opportunity for us overall as society because then we can use these technologies like AI at much larger scale without being too worried about it.