The Banking & Payments Show: AI in payments 2024

On this podcast episode, we explore the deployment of AI in payments in 2024. First, we discuss the most impactful use cases for AI in payments, such as fraud detection and prevention, false declines and purchase conversion, B2B payments, enhanced customer service, and digital wallets. In our "For Argument’s Sake" segment, we rank these use cases in order of impact and debate our decisions. Join the conversation with host Rob Rubin, our analyst David Morris, and Angela Murphy, PhD, client partner at Persistent Systems.

Subscribe to “The Banking & Payments Show” podcast on Apple Podcasts, Spotify, YouTube, Podbean or wherever you listen to podcasts.

Episode Transcript:

Rob Rubin:

Hello everyone and welcome to The Banking & Payment Show, a Behind the Numbers podcast from eMarketer. Today is February 20th, 2024. I'm Rob Rubin, head of business development at eMarketer and your host. Today we're going to discuss what to expect in terms of deployment of AI in payments this year, 2024. For this conversation, I invited our principal analyst in payments and pod show regular David Morris, and Angela Murphy, aka Payments Elsa, and she's also a partner at Persistent Systems. Hi David. Hi Angela.

David Morris:

Hey, Rob.

Angela Murphy:

Hello. Thank you so much for having me.

Rob Rubin:

We're so excited to have you. But Angela, before we jump into it, because this is your first time on the show, let me ask you a few quick fire questions for the audience to get to know you. My first question is: Payments Elsa, what is that?

Angela Murphy:

So that is my... I know there's some giggles. That's real. I am Payments Elsa. You can find me on LinkedIn as that person. And it is kind of a thought leadership identity that I developed being in the space. And then also I sing. I have been asked regularly to sing at industry conferences. So between my expertise and my love for musical theater and my love for Disney in particular, the moniker has just stuck and it's stuck a little bit better than the PhD, if you can imagine.

Rob Rubin:

Well, let's get to that because I noticed that your PhD's in rhetoric, which was a subject I never really loved in college, but how do you go from rhetoric to Payments Elsa?

Angela Murphy:

It's a little bit nonstandard, but at the end of the day, I view myself as a storyteller and someone-

Rob Rubin:

I like that.

Angela Murphy:

... that can bridge the gap between complex concepts and the people that those concepts would serve. So it's really just an exciting space to be in because I see complex technology and I see the people that it can benefit and I stand in the middle as a storyteller, and that's what rhetoric teaches you how to do. It's critical thinking, it's understanding communication strategy, and then it's applying it to the people that matter the most.

Rob Rubin:

Well, now I get it. Thank you. And I hope the audience does too. We have so much to cover. For today I want to talk about the different use cases for AI and payments that is going to have an impact this year. And then for argument's sake, we'll each rank which ones we think will be most impactful this year and discuss our differences if we have any. So let's get started.

Here are five use cases that we believe will be impactful this year. The first one is really around fraud detection, prevention and KYC verification. Angela, why do we think these are going to be super impactful this year?

Angela Murphy:

I think that it gives financial institutions, but also companies and people, the ability to understand who is paying for what and why. I think that with the rise of generative artificial intelligence, we're able to look at the ledger, look at the purchase, look at what it's for, look at the amount and make these associations in near real time that we have lacked the opportunity to do so before. When you're talking about fraud prevention, KYC is a huge boon for banks and for businesses. How do you know your customer? Who is that? Well, artificial intelligence, rather than going through a three-day underwriting process to do the Google search of whether or not their address is real, artificial intelligence can dive in and say that address isn't real, or the Google photos of that address show that place isn't real.

And so I think in terms of where I live, which is the payment space, a lot of folks are talking about faster payments, faster fraud. I actually think that artificial intelligence is going to be the intermediary there.

Rob Rubin:

All right, David, do you agree?

David Morris:

Definitely. We will invariably agree that it's probably the most common application now, and I think there's a need for it. Gen AI I think inherently kind of fits the bill because you're building on top of all the AI-driven innovation that's gone on with fraud. So I think there's a very iterative component there that's probably already being just sweeping across the landscape.

Rob Rubin:

The next use case I want to talk about is related, and it's sort of on the other side of the coin. When I travel a lot, I always have trouble with my credit card because I might use it in New York and then I try to use it in the next place I am and I get a false decline and that really hurts conversion because I might not have time to call and say, "Oh, it's actually me, I'm here." So the next use case we want to talk about is false declines or improving purchase conversion, which is the result of reducing false declines. So what do we think of that for this year, David?

David Morris:

I mean, I'll go in through the purchase conversion window. I think there's some fascinating innovation going on here. If you're familiar with anything that PayPal's released lately, they're one of any number of wallet providers and beyond that are actually coming to market with accelerated offers. For instance, these are generative AI-driven offers that are built on the capacity for a company like PayPal to be able to leverage the immense amount of data at its disposal to be able to really take a look at all of the things that their customer has purchased, but to be able to start literally being more thoughtful about how to predict what their customer may want so that there's a recurring purchase component there to the advanced offer that I think is really powerful. PayPal's done this.

I think ironically, unfortunately for PayPal, it wasn't greeted with the seismic kind of jaw dropping applause that they wanted to despite it being so cool because there's so many other folks already looking at being able to do this too.

Angela Murphy:

Well, I mean, we also know that PayPal isn't really doing their job. Like their earnings calls have fallen short. Their leadership is borderline feckless. Whether or not they engage-

Rob Rubin:

I love It. You're coming back.

Angela Murphy:

Whether or not they engage with a disruptive technology, like artificial intelligence, isn't going to save their earnings calls when they're not demonstrating thought leadership in a space that needs it. They could have been out ahead of this a year ago and they just decided not to be.

David Morris:

I agree with they're being a little behind the curve there for sure. I think this is where they're looking to catch up. But to your point, Angela, one of the reasons I think that this purchase conversion theme is going to really push the envelope in 2024 is that you've got a host of other entities, Shopify and the like being able to do the same thing and harnessing commerce enablement.

Rob Rubin:

See, I'm skeptical.

Angela Murphy:

Well, I'm thinking in terms of commerce enablement and that's a forgotten story for PayPal, I think, in the way that they've approached the market. But that's not today's conversation. We're talking about artificial intelligence. I believe that they're behind the curve and had the ability to be ahead of it and they decided that they wanted to take a sleeping bag to the Hamptons instead of doing something better for their customer base.

Rob Rubin:

All right, I want to jump to the next one so that we can keep it going because I don't want to bash PayPal too bad here.

Angela Murphy:

I love them. You're great. You're wonderful. Just do better on the next earnings call.

Rob Rubin:

Yeah. What about B2B payments as something that we're going to see some impactful use cases for this year with AI?

Angela Murphy:

So this is somewhere that I live. I came into this market through the lens of B2B payments. So businesses paying other businesses and there's a lot of heavy lift and paper lift that is associated with executing a payment from one business to another business. That could be an SMB, which occupies 68% of most business-to-business transactions in the United States economy. It could be something that is more corporate where you think of that top five to 10% of companies that are still enacting these manual processes. So I think that generative AI and artificial intelligence broadly has an opportunity to step in here, solve for some manual processes and give folks in the B2B payment space a little bit of breathing room.

David Morris:

All right. Yeah, well put. I agree, Angela. I think too, and this is something that in real life it's so foreign to me, just this concept of coding. Almost took a coding class when I was 12. I guess I should have.

Rob Rubin:

Pascal.

David Morris:

Yeah. The idea that ultimately AI becomes a part of coding itself. You think of all these highly paid coders out there responsible for all the computer programming, and then think about a technology like generative AI that can read it, that can understand it, and they can iterate on it to actually make that code better. And by better, I mean cleaner, just like it does when we use it to summarize all of our reports here, Rob. Did you know I do that? Does it pretty well. And you can ultimately, if you're talking about being a legacy B2B provider. Fiserv, Global Payments, FIS, they get called out all the time for these legacy tech stacks. And I think there's a real opportunity here for them to be able to put this to use to be able to close some of that interoperability and lower time to market that those tech stacks can create for them.

Angela Murphy:

Can we do a separate podcast about that whole topic? Because David, there's so much to unpack there and yes, that's really where I'm seeing this drive to adoption and the drive for innovation is because we understand the legacy providers are throwing people at a problem rather than technology.

Rob Rubin:

Let's jump to one that we talked a lot about last week when we did AI and banking, and that's enhanced customer service. Chatbots providing better support for customers. Do we see that as something that's going to be impactful this year?

David Morris:

I bet Angela and I both think that there's seismic opportunity here. We probably respond to it a little differently. You already talked about Rob and Angela, the idea that human beings can utilize AI in any number of ways to be able to help themselves. So that includes customer service facing folks, it includes sales merchant facing folks that are able to utilize this and ultimately do their jobs better. They can even utilize Genie chatbot of their own. This is something like the AI pin that's out there. You can actually buy this as a consumer if you want to spend 700 bucks and give it a try. You just tap ChatGPT and you're able to receive help that's telegraphed according to the way that that's been programmed for a customer service department or representative.

Flip that on the other side, and then you have this wealth of personalization, customer service driven stuff that Angela, I know you'll want to talk about too when you think about chatbots and what a chatbot can actually do versus what it's been doing for us, which has never really been good enough.

Angela Murphy:

Right. And I think that we're crossing a different threshold, which is who among us really wants to pick up your phone or anything of that entire demeanor and ask a question to a robot that doesn't really get the problem that you're having? They're not dynamic enough, they're not programmed well enough and currently they're not a part of the data lakes that we draw information from for generative AI. That is a key generative AI use case. There is a data lake of the lexicon of human language through which you can engage and then get to the solution that you need. And a lot of times that doesn't require human intervention. We're all tired of hearing abnormally high call volume. No, it's not. You just have call volume because you exist.

Rob Rubin:

You don't have enough people.

Angela Murphy:

It's not abnormally high. It's been abnormally high since I was 17. Who cares? But it's about having a dynamic experience. When we talk about enhanced customer service, we're talking about dynamic financial experiences and that is what our clients are driving towards.

David Morris:

Yeah, that's fascinating. And there's so many ways, Angela, just like you were saying earlier, so many ways to cut this onion. If you just take a look at customer service and we can mash that up with purchase conversion and let's say there's a chatbot that you can ask about product recommendations. Let's say it's a PayPal chatbot, why not? Why not? And that chatbot gets back to you after quickly and immediately scanning a website or data that's far beyond that to be able to kick back detailed answers and then respond with natural language to your point, Angela. And have an actual conversation where those responses are going to be further tailored to why that person is actually engaging with this artificial entity in the first place. And may actually leave that conversation satisfied.

Rob Rubin:

I think you're right. It might not be PayPal though as we've already previously discussed.

David Morris:

I happen to like PayPal a little bit actually.

Rob Rubin:

Let's jump to the last of the five use cases, which is mobile and digital wallets. Will we start to see AI being built into these things for capabilities?

Angela Murphy:

I genuinely think so, and I think this builds on David's previous point about querying for information. At the end of the day, we have a significant population here in the United States that is underserved and underbanked and there are questions in that population. So it's upwards of 60 million people out of 300-plus million people that reside in the United States. That is not a wonderful figure that we should love. And I think that artificial intelligence and these types of engagement can provide people a sense of security and understanding in regards to their own financial behavior. So when you think of a mobile or digital wallet and you think of the folks that might be engaging in those mechanisms, they're outside of traditional financial mechanisms. They're outside of traditional financial institutions.

Imagine if you have a small business owner that is a landscaper that is underbanked and they transact primarily through Venmo and that's how they pay their bills and how they engage financially in the economy. And they have a question about an invoice that was paid and what that means for their tax return because we're getting into tax season. Imagine that person going on their phone into a digital wallet or a mobile wallet and saying, "I have a question about invoice number X." Or, "How many invoices did I execute on this year?" And then they're not doing bookkeeping.

When we talk about mobile and digital wallets and the underbanked and the role of artificial intelligence, we're talking about solving for use cases that have largely been operated in the cash economy, have largely been operated outside of traditional financial models. And then empowering that population to make better decisions and to take better information to traditional financial institutions to say, "Hey, here's my spreadsheet that they generated through generative AI of all the money that came in and out of my business this year. May I please have $5,000 to purchase a new piece of equipment." That for 67% of small businesses in the United States is a game changer.

David Morris:

I love that, Angela. I was actually thinking of it in a different way like when I think about Square or PayPal or a firm, then the two-sided network. How sellers and buyers are increasingly coming together within these two-sided ecosystems and what opportunity these entities that own them have to be able to facilitate and sales. But then there's the other side of how can they ultimately help them be better consumers and even potentially future clients for financial services.

Rob Rubin:

I see there's so many different ways that we could look at this at digital wallets, the way both of you're thinking. Angela, I really loved what you had to say. I don't want to cut anybody off. I just want to go to the next segment.

Angela Murphy:

Of course.

Rob Rubin:

So today, and for argument's sake, we're going to rank each of the use cases in order of impact in 2024 and argue, really discuss nicely our differences and where we're aligned. So what I want to do is just read out the use cases each and get each of us to tell which one we ranked 1, which one we ranked, 3, 4, 5, et cetera. And then let's see where we're the same and let's see where we're different. And we can finish up the show by talking about that. So let's get started. Fraud detection prevention. KYC verification. Angela, what did you give it as impactful?

Angela Murphy:

One.

Rob Rubin:

You gave it a one. David?

David Morris:

I gave it a one.

Rob Rubin:

Ah, I gave it a one too.

David Morris:

Yeah, I had a feeling.

Rob Rubin:

All right. We'll talk about why in a minute.

False declines in purchase conversion. Angela, what'd you give it?

Angela Murphy:

Five.

Rob Rubin:

David?

David Morris:

I gave it a two. Let's rumble, Angela.

Angela Murphy:

Oh, ready? Five.

Rob Rubin:

We will rumble... So I gave it a two as well. Fantastic.

B2B payments. David, give me a number.

David Morris:

I'll go with three there.

Rob Rubin:

Three.

David Morris:

Yeah.

Rob Rubin:

Angela?

Angela Murphy:

Four.

Rob Rubin:

I gave it a four too.

All right, now we're up to enhanced customer service. David?

David Morris:

Purchase conversion, I'd give it a four.

Rob Rubin:

All right. I gave it a two as well. So Angela and I are very aligned.

Mobile digital wallets. David, where's your number because it could be different than-

David Morris:

Well, I've got nowhere to go but down, so I have to give it a five.

Rob Rubin:

Yeah. Angela, you have to give it a three.

Angela Murphy:

I know. I have to give it a three.

Rob Rubin:

I actually saw your rankings before, but I hadn't created mine and I was happy and not just because I wanted to argue a little that we were totally aligned, but let's just quickly talk about some of these fraud detection, prevention, KYC. Why do you think we gave them all a one as the most impactful? Maybe I'll start because I haven't talked so much. I think it's because there's just billions of dollars of potential losses at play. And the thing that sticks to me is that e-commerce continues to grow. People are shopping and buying all kinds of things online, and the more people are doing that, the more sophisticated online criminals will be. So there's increased sort of fraud that's going to happen and there's billions of dollars at play. Is that the reason why you guys gave it a top priority? Is there a different reason?

Angela Murphy:

No, I think that there's blood in the water as it relates to fraud and AI is going to solve for it.

Rob Rubin:

Right.

David Morris:

It's the first and best fit really in terms of I think how Genie can assess data and be able to spot anomalies and I think make a fantastic impact there.

Rob Rubin:

So the next one, enhanced customer service. This is where we had some differences. David, you ranked it pretty low, four, whereas Angela and I gave it the second spot, number two. Why do you feel like it'll be less impactful than something else?

David Morris:

Well, I want to change that because I thought I had those reversed, and yeah, I would definitely go with enhanced customer services too.

Rob Rubin:

Oh my God.

David Morris:

I'm with you guys on that. Sorry.

Angela Murphy:

We convert him. So the way that we're ranking these I think speaks to the payments cultural milieu around how we see AI affecting our space. And we all see it similarly, however, we're afraid to stick our necks out and make a statement because we could be wrong.

David Morris:

I'm happy to be wrong. I think I've been wrong several times on this show. Rob will tell me for sure.

Angela Murphy:

Oh, thank goodness. So can we hang out more? Can we do more podcasts?

Rob Rubin:

I thought I was wrong once, but I made a mistake. I was actually right.

Angela Murphy:

What?

David Morris:

That's how I've always played this game with Rob. Rob's always right and I'm always wrong.

Rob Rubin:

No.

David Morris:

No, I'm just kidding. So go ahead Angela. Say something that's on the edge of the fence then. I'll do the same.

Angela Murphy:

I think that we need to be ready to be wrong. We're payments professionals. It is our job to stand at the front lines on behalf of our clients financial institutions, whomever. Experience the technology, take it back to them, give them some use cases, provide our thought leadership. We don't have to be right every time. I think that we are so caught up in being right that we don't recognize that taking a chance is a good thing for our clients and stick your neck out. Say something that's wrong, and that's who I am in this space. That's who Payments Elsa is. Is do I know everything about artificial intelligence? No, but neither does Sam Altman and he's the one who coded it. So if he doesn't know everything, I don't have to know everything. All I have to do is stand here and do the best here that I can for my clients.

Rob Rubin:

Very good. I want to hit mobile and digital wallets. David, did you change your ranking here?

David Morris:

No, I didn't.

Rob Rubin:

Angela and I put it in the middle and you put it at the bottom. So why are you a little on the fence on that one as far as impactfulness?

David Morris:

I think wallets are definitely where things are happening, and I provided PayPal as an example, but the reason I would put wallets downstream is because ultimately we're looking at e-commerce, which is web-based as well. I think the chatbot fits very well within a mobile wallet. That's where you're going to see uptake. But a lot of what we're talking about I think applies more broadly. So when you think about enhanced customer service and purchase conversion, I'm looking at those more globally. I think mobile wallets are definitely going to push forward with this, but it's just not as impactful because the space is smaller.

Rob Rubin:

What about B2B payments? For me, I ranked it lower because I think that it's more complicated. Because we were doing rank impactfulness for 2024, I pushed it to the bottom.

David Morris:

I think that that's probably where every single mid-size and up B2B company is trying to do something. But to your point, owing to the complexity, owing to I think how difficult it actually is to be able to train these language programs at a high scale, and you're also looking at the compliance issues that a lot of these B2B folks have to jump hoops through, you're going to see more incremental change across that board. That would be what I think.

Rob Rubin:

Angela?

Angela Murphy:

I'm going to be interested to see because B2B is archaic in its behavior and is slow to shift. While artificial intelligence can solve a lot of problems in B2B, I think that your engagement and your uptake and your adoption will be higher in the consumer world because there's two things. One, they're more used to it. The Amazon notification of the individual experience really has reached a critical mass. So for those folks, AI is not going to be that hard of a leap. In their professional life, however, we have seen that the check is still holding on really strong.

David Morris:

Right, write checks.

Angela Murphy:

The companies write checks, but their employees are using Venmo. We are literally looking at an incongruence of commercial behavior as it relates to technology. So what is our role? Literally doing podcasts like this.

Rob Rubin:

Right, exactly. So the last one, we're going to have it out a little bit, but let's see if we can do it in the minute that we have left. False declines, purchase conversions. I think that I and Angela put it at the bottom and David made it to number two. Did you change that? Maybe it's number three now.

David Morris:

I think it's because I actually see kind of the enhanced customer service and purchase conversion linked a little bit. And I think that there's already a lot going on there. I think we provided some examples. So I'll raise my hand for PayPal here again, and I actually think that these kinds of offers on steroids, they're going to be commonplace by the end of the year across wallets because they all have the scale to be able to implement them. And then it's a matter, I think, of being able to pursue that downstream a little bit toward retailers that aren't as big as Walmart because they're going to be doing it too.

Rob Rubin:

Okay, I'm going to do a quick recap of what we said. So on fraud detection, prevention, KYC, we all said number one, that'll be the most impactful this year. Enhanced customer service. Angela and I had it as a two. David started with a four, but we turned him around to a two, so we were all aligned there. Mobile digital wallets, Angela and I, again had it as a number three, whereas David dropped it down a little bit. I think he dropped it to a number five. And B2B payments, David had it as a three. Angela and I had it as a four, but I think we all agree that it's just taking businesses longer because of the complexity and their slowness to adopt new technologies. And then false declines in purchase conversions, Angela and I had that down at the bottom, whereas David pushed it up top, but I'm sort of understanding his rationale around using it for purchase conversion and enhancing customer experiences. I just want to say this was a really fun show to do today, and Angela, I really hope that we can have you back a lot.

Angela Murphy:

I would love that. You can have me back anytime.

Rob Rubin:

Thank you. It was a lot of fun and laughs and I learned a lot just listening to both of you. And David, you know I love you and you're always on the show and I love that. I want to thank everyone for listening to The Banking & Payment Show, an eMarketer podcast. Also, thank you to our editor, Todd. Our next episode is on March 5th, so be sure to check it out. See you then. Bye David. Bye Angela.

David Morris:

Take care, guys. Thanks a lot.

Angela Murphy:

Bye. Thank you.

Rob Rubin:

Awesome. Thank you.