Fraud Trends, AI Threats, and Consumer Trust

We throw our consumer base and have them transact your story. Oh, good. We're going to take it all away. So we find this, and by the way, those that want to take it all away, they're not bound by regulations or sign -up of passing, and they really don't care how much version they create.

The more version they create, the wider they pass their net, the more they get back. So they don't face the same limitations as we do, and we title this presently based on the more and more faster world, because if you stop and think about it, our landscape has changed in the ways that we've done money more in the last five years and in the last 50.

And so we've made a very complex environment that just creates a lot of new vectors that can be compromised by the product. But your consumer behavior, your human behavior doesn't change. In fact, you want things fast with no friction.

I think about Willy Wonka and what was her name, Violet, that was like, I want it now. They want it now, and they don't want it to be halted. So when you think about it, in some of the conversation this morning, I mean, it's AI, AI, AI in this presentation.

And I won't have a whole presentation to be about AI, but it's a broad presentation. I think it's the most applicable place to talk about AI. But it's really an AI arms race. Coupled with human behavior and the macro environment that's changed so much, and all of us compound, and everything's moving faster.

And so we're going to talk about how combat is brought in the past world.

So I talked a bit about some more friction. And from the consumer's perspective, if they experience a false decline, it means that you don't know them. And I think that we strive to really make sure that our consumers, our car holders, our members, our credit unions, that you know them.

That you know them well. One measurement that we use to track friction is how many false positives do we have? How many good consumers do we impact at the point of sale, making a good transaction, to catch one across?

Product moments are really trust moments. And trust is your currency.

We world that currency or that trust by doing well in this regard. And I've met many of you, probably most of you, but for those that haven't met, I'll pause and introduce myself. For whatever you've done yet. And I lead our fraud space here in North America, our fraud detection space primarily, as well as credit risk.

So many of you are adopting tools like product manager, advanced defense. Those are supported by my team.

So, we also lead a team that performs the house service product management for about 3 ,000 beauty banks in credit union. And we take it very personally and really that this is our job, is to grow their currency.

And I agree with seeing that it's worth the job.

When we don't do it, or when you don't do it, the impact is real.

We did a study on our own data. This has been a little bit ago. But we found that in a six month time period, one in six consumers had experience of false decline. When I think about that was just six months.

But really, show of hands, how many of you have been in five to a point of sale with a bad decline?

And how did that make you feel?

You had a motion around it? Bad. I know, uh, sad.

It happens to me. I feel judged. I feel delinquent. I look around. We've seen this happen. Someone's saying to me, ma 'am, do you have another way to pay for this? Well, you don't have one way to pay for this.

This should work though.

And I feel like, I have to defend myself, right? It's not a good emotion. And we don't want our consumers to have that emotion. Because 20 % of them will stop using it. And if you do it repeatedly, they'll walk away.

And so, it's very important that we are very meticulous in how we make our decisions.

So, we're going to dive into artificial intelligence for good and bad. We'll talk a bit about the human behavior as well. I've got some scenarios to walk through. And how do these combine and create our new reality?

So, I'm from here. Home on Nebraska. Welcome. We also have Warren Buffett.

That's about what we have. This is called World Series in Warren. So, we have the best of the nation's best do, right? The world.

The world. The world.

So, if you have time to do that. I know you do. It is

amazing.

Yes. I'm sorry. I'm not going to do that. First, let's talk about AI.

I was at, I saw some of you at an event a couple weeks ago that he's at New York and there was a great professor from Carnegie Mellon and gave a very good example that I really liked to help us just think about AI.

And she said AI is like a fork in a knife. And we all know what the fork in a knife were designed to do.

But if the fork in the knife are at the hands of the wrong person, you can poke an eye out. But if it's the right person, you're nourishing your body, you know, eating your soul.

So, if you have a beating, you're hearing about it. So, when you think about the threat actors, it's about synthetic identity creation. It's about the deep, big voice hands that you hear about. If you haven't, you're getting to have a code word.

My mother knows the code word. If my daughter ever calls her frantically needing money because she's in trouble, what is a code word? And if she can't say it, then it's not my daughter. Someone has made it sound like my daughter, which is horrible, right?

I guess it would think that someone would do that, but they do it at scale.

It's no longer a Nigerian prince coming to marry you. You need to get there, right? But the written word, we don't have grammar issues. It's very elegant. I use it myself. I use coaching to make sure I pick up the column in my email sometimes.

And that was one of the best features. Instead of saying, like I said before,

I'd say, maybe you should instead say it.

That's a dispensary. That's a dispensary.

That's my last message.

And they can scale it, right? So, it's not like run and done and they're operating. So, they can scale their efforts, no walk or any channel or stop. But for good. We've talked a lot about agenda conference this morning.

I have so many busts and things on that. I have a different presentation on it, but I won't go into that here. But it does create amazing member experiences and shocking. If I can say, I want these in the 90s.

I want them by December 15th. And when they go below $100, you know, for the speed level even, go ahead and transact and make that happen.

Today, we talked a lot about this morning. It's really about comparison shopping. And then it sends the user on a journey that says, here, go to this website and finish the transaction. But it's really not on time. But it will be.

And we had great conversations at lunch about the bidding war that happened on which card is going to be the best for use. And so, thank you for that conversation at lunch.

We look at it for faster fraud detection, reducing that friction. We're doing that with POC right now with IBM. The managed service group that we have where we do this fraudification for 3 ,000 FI. when we look at adjunct analytics.

Other three are successful there. Then we think about going it out to OCTO as well.

And then ease in customer resolution. We think about these few areas of the best place to help adjunct processes.

Now we'll talk about Warren. So Warren Buffett saw an AI representation of himself. And he said, based on the one I saw recently, I practically sent money to myself.

And he's a pretty recognizable person. You know, think about that for our families and our friends and how they can be easily viewed.

We'll hear more with Warren later.

So we're going to jump into the human factor. And there's really two ways that I'm breaking this down in social engineering and first -party fraud. But before I jump into those, I think it's important to look at this because I told someone I would say this word in my presentation.

I mean, do you play whack? Right? We are constantly trying to find the next way that the criminal is perpetrating fraud and we knock it down from a tech perspective. Well, modern fraud is not looking to break technology.

It's looking to break people and process it. Because we've gotten technology to prove it. And so now it's about the human factor. So first we're going to talk about man in the middle being him. So who here has been impacted by a children provisioning fraud?

Anyone who doesn't want to raise a hand probably will. Many of you have. And I want to talk a bit about how that's perpetrated. I think even as consumers just for a little bit of education. So it's really when they're intercepting the once -time passcode to add the legitimate card into the criminal's wallet.

So this example walks through.

Who's gotten a DMV or a toll text message? That's a good thing, right? And here's one of my team. This is the legit one that he had to change in Nebraska DMV. And it's saying he has a parking ticket.

In my house. This could be legit because of my email. This is on my daughter's title. And she's 25 and she parks in randomly. It's just an option. She gets parking tickets. So we always call her first in case she knows something.

And she goes to the legit site to see does she know parking tickets. But let's just say for this example that I'm a good, um, a good consumer, a good, um, resident. And I want to make sure that I'm making this payment.

So I click the link. What happened?

Well, it wants me to fill in my details on this payment portal, right? It's not a legit payment portal, but it says, please fill in your details. So I give it my card number, my expiration date, and my date, my CD2.

And what happens is that is, um, the criminal screen pops up here. The dark screen. They are trying to use that data and information to provision their, your card into their wallet. And so this is their attempt to provision into their wallet.

And they get stumped with an OTP request. They have to step up authentication.

So what are they going to do to get that OTP? Well, they already have the citizen in the payment portal. So they say, well, I need your, your, to complete this transaction, I'm going to send you an OTP and use your OTP.

The OTP is coming from their bank because the bank thinks that they're provisioning into their wallet. And they're giving them payment for a card. They get the OTP. They enter the OTP. And now the criminal has successfully, um, loaded their card into the wallet of the criminal.

And they don't do this. Not one guy, someone, one gal, someone, somewhere. They do it. It's still, this is an example from the telegram, um, you know, a phone bank. Ten phones in there with four of these cards on, on each phone bank.

Right? And they're ready to then go over to create fraud with these or to sell them to others. It's a whole, just as we sit here today, they collaborate. They are collaborating and there are vendors and there are testers and there are, um, sophisticated novels that people are writing out.

So they all get together and they figure out how to novelize this and scale it. So that's just one example of ways in which social engineering can happen.

Another, we dubbed this the Scandemic. Because during the pandemic, Scand proliferated. Scand are not new. But, um, during the pandemic, we saw mass adoption of e -commerce. We saw debit outpost credit. We used online. And we had a lot of, you all call digital doobies online that aren't maybe as savvy.

I mean, I can name some people in my family that I worry about. But we all went there to do health, to do work through jogging, stream things, workout, you may be. And so, um, in this stat from Javelin, three in ten consumers report being in the Scandemic at least once in their lifetime.

Just pretty staggering. So back to Warren.

He said, he said, it was not this year, but it was last year in Albuquerque Gold's annual shareholders meeting. And in last year's meeting, he said, scamming has always been a part of the American team. But if I was interested in investing in it, it's going to be the growth industry of all time.

And he was a good client in this book.

So we were interested to understand the growth of scams compared to the growth of other well -known trademarks that we are all familiar with. And from 2018 to 24, sometimes for our data's stay a little bit, the reports come out.

But, um, scams grew 845 % compared to the next highest growing, which was Tesla at around, you know, 510%. Google, Amazon, Microsoft, and Apple, you can see their growth. So it is one of the highest growing things of all time.

But it's probably still looked at that back in 2019. And this is just for how they get reported to the government. Anyone want to guess how much fraud actually gets reported to the FTC?

You can't hear very well. 40%. 40%. Anyone else?

50%. 3%.

It's around 15 to 20%. So imagine if all fraud were reported, this growth, and what that would look like. We only see a few of 20. And who would guess what the top reason is that people don't report it?

Embarrassing. Embarrassing.

I thought so too. It's a very valid, strong answer. And we were at a FIFA event, and there's a report coming out a couple weeks, but I was good at it, and I didn't really think about it, right?

Where the Scam Alliance did a report with Parker. And they surveyed 5 ,000 US consumers to find out about scan reporting and maybe why they didn't. The number one reason, they didn't think it would do any good.

They wouldn't help anybody else. And the number three reason is they don't know how. I honestly don't know that I sitting here today could even say that most people know how to go and report the government, right? We can probably check the VTA and get those pretty fast.

But then, in change.

And then, we're going to look at the data and work to understand. We can't fully action. So, there'll be a lot more on that.

So,

we then look at the data and work to understand. We always think about our values, right? So, we always think this is always happening to our parents and grandparents. And that's about the younger generations, which we're talking about later in this conversation.

But we looked at the losses that were reported by those 60 and over, where the amount lost was more than 100 ,000. So, only when it's more than 100 ,000.

2020 was 55 million. Only when it's over 100 ,000 for the incident and only over 60.

2024, it grew 700 % in that same scenario. Just to kind of help also understand the magnitude of this data.

So, switching gears to first party fraud, the second human factor.

It represents an estimated one -third of all fraud claims.

It has more than 100 % increase when you compare 23 to 24. And it has become the most prevalent type of fraud worldwide. Sometimes it's hard to buy. Sometimes it's hidden in the collection. But these are estimates based on a famous journal report.

So, I'm next going to walk through what's happening in society that might make people gravitate and become first party fraud users. First, we're going to look in the mirror, degrading moral fabric. I'm sure none of us are this way, but we're going to walk through it.

And then some online gambling conversation and then economic pressures and how they impact it. So, first, I always think it's fun to immediately get this after I go through this, but I always joke that I'm going to put all the German people in this corner and all the millennials in this corner and gave them the words back there and then we all judge each other.

But this was a SOFIA study. And we just pulled out some of the snippets and drove through some conclusions.

19 % of Gen Z do not consider first party fraud to be ethically wrong.

And almost 20%, one in five, don't feel that first party fraud is ethically wrong.

That's three times the rate of the baby boomers in this survey. Now, I'm not saying the baby boomers didn't feel that way when they were younger, because there is an age, you know, a generational difference, but I think that we've seen, this is a whole different environment that we're operating in today.

There's a lot of reasons, socially and economically, why it might be different.

30 % of Gen Z admit to making a purchase through a BMPL loan, but they don't ever intend to pay it back.

52 % of Gen Zers would commit first party fraud at government.

So if you think about a couple of these data points, 33 % know it's ethically wrong, but they would do it anyway. Right? If you think the 52 and the 19 kind of meet in the middle, it's ethically wrong, but they do it anyway.

The second is online gambling.

We're now at, I should have checked this, but I think we're around 40 states that it's legal to gamble online.

It's expected to double, online gambling is expected to double by 2030. And there's some estimates that account for online gambling fraud losses to be about 15 % of the total transaction, but 55 % of the total draw. So first party fraud losses, 55 % would be attributed to online gambling, but it really represents 15 % of the transactions.

Obviously, there's obviously higher risk in an online gambling transaction and allowing those things to happen.

So why is online gambling trending? We joked a little bit this morning, we joked about betting. Yeah, I think you've been a joked in years about, and we joked about betting, about the market and how people are getting in trouble betting.

I think that's a little bit more money, but 43 % say legal sports betting is a bad thing for society.

58 % of the 18 to 22 year olds engaged in at least one sporting activity in the last year. This is just meant to show generational differences and how we have different mindsets.

A lot of people aren't going to agree with it. 58 % of the 18 to 22 year olds did at least one time in the last year. And when you get down to the students that live on campus, 67 % tend to bet at higher frequency.

They are better than they tend to bet at higher frequency.

And it's changing. So we now have prediction markets. Polly market, Cal Sheehy, Ronald McDonald predicted.

These markets are legal for 18 and over in all states. So it was common gambling for these states. Now we've gone to gambling on predictions. And it's legal everywhere.

You gamble on things like, we've been in the news a lot. But you can gamble on things like, will Zelensky wear a suit before July? And then there were pictures of Zelensky wearing what some thought might be a suit.

So then there was a lot of argument about, was that a suit? Or was that not a suit? And what do you do if you think that wasn't a suit and you lost that bet? It's beautiful. Right? So it just creates a lot of volatility in this space.

And I don't think it's going to change because Donald Trump Jr. is a strategic advisor for a calling market and now for Cal Sheen. So I'm not being political or making statements. I'm just saying if Donald Trump Jr.

is an advisor to two of these, I don't think we're going to see a lot of change or regulation around it.

This just shows the stats, the winners and the losers for the Lillel Zelensky wear a suit before July. It's fuzzy, but those are hundreds of thousands of dollars by the sources. Those are individual, the winners of bars. And it shows that a big scenario is the losers.

Show me that song of the people.

We won't do that.

All right.

DraftKings and FanDuel recognize that calling market and Couchier are entering their markets. And so now they're entering prediction markets. So it's just growing and then deliberating.

So we're going to talk about generational differences. And I think they'll get these slides later, right? Definitely, yeah, you'll get these slides. But I just thought it interesting to think about from boomers to Gen Z, they all had different insurances in their life, which form their intrinsic nature.

They trust differently. They can be scammed differently. And there's the proper and best ways to educate each generation. I provided this presentation to a group of friends a week or two ago. And they're trying to grow their younger member weight.

That's our many of you. So I think it's important to think about ways to educate them and how they may be scammed. And then at the bottom, tolerance for friction and then preferred payment channels. So I'm wondering this slide, but I would say generally if you think about it this way, we all trust differently.

And boomers can over trust authority just based on their trust and what they grew up with. And Gen Z can over trust authenticity. And scammers know this. So while we think that boomers are at higher risk, they're not actually Gen Z's at higher risk because they over trust authenticity.

So LW is the real, but I think that scammers and threat actors understand this. And they need them where they are and where they're most vulnerable.

Finally, in the first party fraud front, economic pressure increases. So what's probably not to do with price bait? Maybe that's a sale, I don't know, but a good fall tool in price. But actually quite a lot. So we'll talk a bit about the growing big economy.

But if you think about the big economy has grown. Uber, Dash, Uber, DoorDash, those that are taking on work. Especially when they have government shutdowns and they're picking up work like that so that they can pay their bills.

It's a very volatile income. It's not stable income. Consumer debt is an all -time high. And when we see these things happen, sometimes good people will make decisions that they don't have to, but they feel compelled to in order to respect their funds for what they treat the need in life.

36 % of the U .S. reports engage in some form of big work as of 24. And it was projected to rise to over 50 % by next year. And we're at 45 % right now. So that's pretty much on the contract.

49 % of Americans have used, or that have used BNPL. 41 % makes a base -run payment. And 25 % use it to buy groceries. Now I don't think that the average is going to be able to be used for groceries.

I think it's a big ticket item. But it's available to them as a service. It's kind of like back in the day, you'd go to have a price store and lay over your Christmas gifts for the year. It's like that in the lab.

I think it's more for families, but I think it's better than you get about it. But it's definitely a problem. And then 90 % of the present in the holiday season decided economic hardship has promoted. When you think about a dispute workflow, you ask a consumer, did you buy that?

And they're in an economic hardship situation. The odds of them saying, no, I didn't, probably go up.

So, I don't always mean to like spare, and I have other content that kind of talks about best practices for the transaction. And I'm happy to talk with any of you about any of this at Bryce. But I think that I wanted to leave you with a few questions to take back.

So, the right question is not always how much fraud did we stop. It's how did we make that consumer feel that it happened back to our friction conversation. You all have a risk tolerance level. Some are greater than others.

You each operate different types of portfolios. And you all have a budget line item for fraud losses. And you forecast for it, hopefully. But I think thinking about how a consumer reacted to false decline, it's great to retain them and retain that interchange and retain that relationship.

And sometimes a higher fraud loss could be violent for that. So, I want you to think about what does fraud friction cost you in loss, spend, and nutrition?

As you work to attract younger generations, are you thinking about those behavior differences? There's a lot of other behavior differences. I just highlighted some of them. Are we protecting card holders or metrics? Sometimes we get proud of our metrics and we don't think about the ramifications or the other parts of the business that an authorization decision can make.

And then we have test over digital loans. How do you have fraud experiences compared to other digital loans? Are we thinking about that end consumer and how we're digitally communicating with them or interacting with them? How do you have fraud experiences compared to that?

So, I have some time. I guess I could have added other content. So, I have some time. We want to open it up for questions before we welcome Dan and Mike Tucker.

I know I'm not talking. I know that sometimes when I speak, I have to remind myself it's not all the fraud professionals from your institutions in the room. We host our risk counsel every year in November. We'll have it here at the Marriott next door.

I'm going to ask you for early November. So, I encourage your teams, your broad professionals and your teams to come and engage with that. Many of them do. We've done that for a long time. And then, I'm going to probably get 20 questions.

And a lot of conversation happens at those events. So, just a little plug. We'll be sending that out for the break of November down here at El Hock for real time.

Question over here? Mine.

We

started with the, like, I know the event of the commerce transaction is really very low value right now. Erica went down, she saw like 20 from that. Have we started this in fraud in any of our, like, maybe office, like actual fraud occurring that's coming from the agency?

Yeah,

I

would say Erica mentioned 20. It's probably our own people buying on Etsy. And we're testing the flow. And we've done that from a IHOP with the team with that just to see what the experience is like. I would say no, because we're not seeing truly autonomous agent transactions.

We're seeing consumers being led by their chats to the right website to make that purchase. But it's just a legitimate purchase e -commerce. So, no, but we're writing a lot of data elements in our system that can recognize.

And we're adding more. So, we're definitely prepared for it.

So, I want to move on to maybe a lot of risk counsel, fraud counsel meeting, but the dance has been sort of who you guys are doing.

Got any help of my crazy woman? So, the main expense has been really where you guys have supported our position to sort of straighten the AI. What do you see as the next immediate evolution of that, either an advanced expense or an individual?

Yeah. Thank you, Mosa. So, many of you are so keen to come in and use advanced expense. We brought that to market in 2020 a lot. We have about 70 % of our accounts go for advanced expense versus the benefit.

And it has a component of AI. IBM would call it general AI. Everyone argues about the definition of AI and general AI, but it does work to source the broad data that we have in the environment to find the best rule with the right threshold.

I thought that that would be the number one tool used by everyone who adopted advanced defense, which we're going to use. It does have limitations. You can only have someone's data go through it and so many data points, but it's there and available for use.

We also have an advanced defense, the way to upload custom models, which many are taking advantage of that. So, we talked earlier about it. You asked earlier about AI into the FISER systems. I know it's broader than this broad, but we do have a lot of success with several of our issuers firing our own models in advanced defense.

But your question about what's next, we started that business in 2018 and brought into the tournament in 2020, and it's 2022.

And we hear very positive things. We actually have two advanced defense defense group sessions now, and they've been very positive. We get a lot of great feedback. But we are already thinking about what's next. And I've been chatting to David about this as well.

I think that, and I'll talk about it a bit in the panel conversation at the end of today's event. But I think about every event should be fed to a system that has AI at its core and can find the product real time using AI.

I shouldn't have to. I could always have to rule venture, but I want to live in a world, and we will, we take this downright, but where every event, not even just part transactions, but IVR data, customer service, like wherever you're engaging your consumer, if we had everything that was happening fed into one engine, and if it real time made decisions based on all of that data, that's what I think about.

And honestly, most of the time we think about who's the best provider to partner with to do that, and every day there's a new entrant. Because don't we just have to be a fraud vendor or provider to have that type of system and solution.

And so we're actively thinking about that right now, and what does come next. And I mentioned earlier, we service and support 3 ,000 community banks and financial institutions on a whole different system, on a whole different platform here in North America.

I want to have that and all of you in one environment. Think about that conversation. You know, we talked in our first speaker today about the value of data at scale and conversion. I think that's where we need to get to.

And so, hopefully that's a goal of mine. Good job. Charlotte, we have two goals. Yep. Hi. Do you know if issuers are reaching out to customers who experience the false positive decline and doing something for them or engaging with them?

Yeah. So, I mean, if often when you decline, you queue. So, often if you do make that decline decision, and unless you know it's brute force attack or you don't want to further cross -striction, you just want to decline and knock it down, the consumer never needs to know.

But if you're not confident that it was a strong decision, then typically you queue and then you reach out with a digital communication or a call to say, hey, just checking in. We just impacted your ability. Was that your ability to transact or did we get that right?

And then after that, then going through the whole cutting part down and creating the plastic. So, absolutely.

Anybody else?

Alright, well then I left these two fine gentlemen a little bit of extra time. You're welcome.

So, if I'm going to see you in the end of the century, I'm going to talk about their two great connections.