The Exchanges

Q Okay, so you are a believer then in this vision, and it sort of leads me to a natural follow-up, which is do the chatbots sort of become the storefronts of tomorrow? Are they going to be the place where we do most of our transacting? How does a chatbot fit into the picture of, you know, transacting online?

A Uh, it's, it's, the, the way we think about it, Alex, is, you know, before online commerce existed, um, you know, you walked into an AMC theater to buy a movie ticket. Then you decided to buy sometimes online. Then you sometimes decided to buy it because you clicked on an ad on a social media website through their app. And that doesn't mean that physical commerce went away. It just, they were like, The share shifted toward more channels in the way human beings actually interacted with it. And we think with the Gentic, you're just going to see another channel that's going to be very helpful in certain types of commerce, in certain segments of commerce, um, which doesn't mean that, you know, the existing channels are going to go away. I just think there's going to be a good amount of migration from how things work today to a portion of that is going to move to a Gentic. So like, that's, that's, that's, that's, that's the hypothesis we are working with.

AI assessment note: “you're just going to see another channel that's going to be very helpful”

Q the discovery process could look like, because this idea that, you know, we might provide the LLMs with lots of information about us. I think we're already there. I mean, we probably share more with LLMs than anything else. And we're being honest, some of us are in love with these things, not me, but some people. So how could this discovery process, uh, look as, as Visa gets involved?

A Um, I think the simplest way to think about it is You know, you use your cord to do a number of things in life, and a lot of those are frankly not used at LLMs, right? You walk down the street to like Pete's or Starbucks coffee, and you like to stay at the Ritz-Carlton versus like a Motel Six, and you typically like to fly United versus Spirit Airlines. Now, those are just things that while you're Agentec platform surface area, LLM, may have a lot of information about you. Those are just, like, personal preferences tied to your long-term spending preferences. And what we believe is, if the data is private, if it has full consumer consent, if it is totally tokenized and no raw data has actually moved anywhere, but can there be a bunch of signals that, you know, you, compared to like other spenders, That a agent or a company can sort of like consume based after you've like fully consented to it. And after consuming, say you say, I want to go to Tokyo. I don't want to live in the, I want to stay in the Ginza district in Tokyo for like a week. Um, you know, and give me hotel recommendations. Now, knowing that you like to stay at the Ritz-Carlton or knowing that you generally have a propensity to stay in, like a signal that tells that you have a propensity to actually stay in a Luxury hotel would be very helpful for a LLM to say, oh, you know what, here are the options which we thin…

AI assessment note: “a signal that tells that you have a propensity to actually stay in a Luxury hotel”

Q How long do you think it's going to be until you have that data? Um, I guess that's a way of asking like, how, when is this going to be real?

A Look, I think there's, I think you're seeing, starting to see versions of it become real, but it's again, very early days and these tend to be like, you know, very, very specific transactions. You're starting to see, you know, some of this go live in, we're already seeing some form of agenda transactions in different ways, shapes and forms with like various AI platforms. But again, it's very early. It's like at limited merchants. It's like, you know, in beta is largely in the United States. I I'd say in about six months or so, we'll start to see like some early statistically relevant enough data sets. Where we can start to like draw conclusions. That's like my best guess, but you know, it all remains to be seen how quickly some of these things get adopted.

AI assessment note: “I'd say in about six months or so, we'll start to see”

Q messaging app, You're effectively, you have a single source of your conversation. So they know who you are. They have your payment information stored. Um, with LLMs, it's even more advanced, right? It's like they know who you are. They have your payment information, uh, probably, and they have a deep understanding of your interests, your preference, sometimes who's in your life. So the potential there to me seems vast.

A Indeed. And I think what's changed from the example you pointed correctly pointed out is now Take all of that personalization and context, and then you add how the LLMs and the agents interact with the outside world and their understanding of the outside world, and then connecting those back to your preferences, which is where you start to really, really connect the dots. So while you may have had a payment credential stored in a messaging app, for example, and you know, we've seen wallets in parts of Asia, Latin America do this very, very well. Ultimately, It was still, while it was intelligent, it's still dependent on sellers and, like, onboarding themselves onto these platforms to make themselves available, um, to consumers with their preferences. And I think in the new agentic world, I think that becomes an even easier sort of, like, connection to make. So I think what he's saying is accurate, and I think it's going to become even more real with how quickly we're seeing How quickly we are seeing this genetic bots like be so useful in things that human beings would ordinarily spend a lot of time doing.

AI assessment note: “Indeed. And I think what's changed from the example you pointed correctly pointed out”

Q Now there is some work on the back end here, um, to make this stuff work more seamlessly. For instance, uh, Visa's working on something called the trusted agent protocol. Can you tell us a little bit more about that?

A Yeah. Um, I'm, I'm happy to. So one thing that we started to realize is as and when agents have payment credentials associated with them to make purchases, you know, you're going to see some really, really good applications of that in, you know, some of the things that, you know, some open AI and others have released, like ACP is a good example. AP two is another very good example. And, you know, we're working with them to figure out like, how do we make sure that cards are safe and tokenized and all of that, but ultimately, you know, We still have, we still have to make sure that there are like two to three hundred million merchants around the world in nearly 200 markets, and ultimately agents are going to go have to transact at them on behalf of consumers with depending on, you know, you can debate how much level of autonomy they're going to have, and we want to make sure that the existing web infrastructure that That works today for e-commerce also works for an agent e-commerce world. But in order for that to happen, there are different problems that have arisen. Now remember, in the regular e-commerce world, bots were not a good thing, right? For the last 30 years, like, bots meant bad things. Yeah, I mean, you know, the e-commerce industry have spent a lot of time trying to keep bots out because typically it meant that someone was actually trying to get in and do bad thing…

AI assessment note: “we want to make sure that the existing web infrastructure... also works for an agent e-commerce world”

Q things. Um, what, what is the visa approach going to be if, um, let's say my agent goes out and uses my visa card and buys a lot of stuff for me. That I didn't want. You mentioned you can trust Visa. You go anywhere, use Visa. Visa's got your back. Does it have my back if an agent goes out and buys some stuff that I really didn't want?

A Um, look, that's, that's the fundamental question we spend an enormous amount of time, um, fixing and thinking about. And again, like I can give you a parallel, right? What does that really, really look like in a e-commerce world? That's not agentic yet. You know, someone can actually take your card and actually use it. You know, PII and PAN data gets leaked all the time. You know, hackers have access to like credit card information. Stuff always happens. And so everything that we design and build, we are deeply, deeply focused on how do we make the, make sure that the vectors and the surface area of fraud is as minimal as possible. So in the Visa Intelligent Commerce APIs that we introduced earlier this year, and some of the work that we're doing with our partners, we are uber focused on ensuring that any Visa credential That is used in an agentic environment has a high level of trust and security associated with that. So let me explain what that means. So if, say if you have a Chase Sapphire card that you're using for a specific, you know, agentic app on your iPhone. Now, if we can tie your root of identity to that app on that device, And we can assign a token where if someone actually tried to use it anywhere else, not like that's not you or not in that app or not on the device, the transaction just wouldn't work. So we actually upfront are working very, very hard to create …

AI assessment note: “we are deeply, deeply focused on how do we make the, make sure that the vectors”