Modern fraud solutions for modern fraud problems, with Rob Meakin (Creditinfo)

You know, you can put as many controls that you want into combat fraud at the front end, but the consumers become completely intolerant of any friction in those interactions that they have with financial services companies. And that becomes very critical when you're entering into a new relationship with a consumer. You know, you've got one chance to make a good impression there. But we're really trying to do, since the three simple questions for clients, is this presented identity or genuine identity? Am I interacting with the owner of identity? Are there any risk factors associated with the identity I've been presented with? My first job involves building fraud detection models and so whenever I hear about innovation in that space, I'm interested. My second job involved working in Kenya. And so whenever I hear about innovation in that market, I'm interested. In today's episode, I'm talking to Rob Meekin from the Credit Info Group about fraud detection models in Kenya. And so of course, I was doubly interested to hear what he had to say. Right, well, Rob Meekin, director for fraud and identity at the Credit Info Group. Welcome to the show. Rob, your entry into financial services is different to most. I studied chemistry at university and that was my love all the way through school really. And you know, coming out of university, teaching was an opportunity to travel in some way. So I did the teacher training and then to spare off degrees for a few years, which was a great experience. And then came back to the UK and started teaching high school chemistry. To be fair, it's probably a different proposition entirely to the little kids out in Greece teaching the maths, they've been on it. And you know, that teaching career here came to an end fairly promptly and then was taking an operational role in the call sensor and with British guests and working up from there. That 10, 15 years I spent in the utility sector really gave me a good grounding in terms of understanding what problems face businesses in terms of operations running back off his processes and how that sort of interaction with the front end consumer experience was absolutely critical. So yeah, as I moved through those operational roles into process re-engineering roles, into more technical roles and then into financial roles came out of sector really having managed to bad debt P&L for British gas and seeing how that really made a massive impact on the profitability, you know, hundreds of millions in terms of write-off. And you know, once somebody gets into debt, it's very little you can do about it. So you know, that spawned us to think about, well, how do we use the data that we can get hold off from, say, credit references alongside of the demographic data and apply models to drive inside out of that data so we can make more upfront decisions in the same way as, you know, financial services do, but actually applying it in that utility space to help consumer sort of like not get into that in the first place by recognising that risk early on. Sure, you can heat your house to 25 degrees with the Windows open or something, but realistically, it's not like a credit card where people run wild and buy all sorts of things they probably shouldn't buy and fall into debt. These are us going about our daily lives. And so an interesting place, I guess, to cut your teeth and sort of see firsthand what doesn't, doesn't work. You know, I think that was a great grounding then for moving across into B2B roles with originally called credit, who were then subsequently acquired by a Trans Union, focusing really in the product space and building up solutions then to help clients make more informed decisions and, you know, on the other side of the fence and where I was in the utility sector. You know, and if you look, say now it was more specific if you're around the fraud challenges that organisations face, you know, the cost is significant and it's not just costs in terms of forward losses. You know, the actual losses, I've seen estimates saying that the actual cost is four times at least the cost of the fraud loss. You know, you've got those back off this operations, you've got the cost of all the anti-fraud controls and the vendors like Credit Info to pay for that. And then, you know, I think probably the thing that escapes, escapes people sometimes in the fraud spaces that, you know, you can put as many controls that she wants in to combat fraud at the front end, but CX is king. You know, you put too much into that front end process, introduces too much friction, increases abandonment in your sales process and then you've got a hit to your top line that represents probably a great cost in the new forward losses. Yeah, it's a tricky thing to get right. Also, well, I'm busy going through the startup process with a couple of businesses and, you know, investors talk about asymmetrical returns, you know, where can I get extra bonus returns? And for me, when I look at fraud, it's like that, but on the negative side that bad debt to some extent, you can predict and control. But if you miss a fraud, if you don't see a loophole in your system, it doesn't really matter what your other controls are. You can lose a massive amount of money very quickly. It's not that one customer's borrowing from me. I'm giving them, you know, $500, $1,000 credit limit and they might spend it all and not be able to pay us back. Is that one fraud syndicate might take out a thousand loans and makes them out? And that's what makes it quite frightening. And they will talk about what that world does look like today. You, with credit info, got a bird's eye view, but not just of one market, really other, a number of very diverse markets around the world. So as you look at the state of play today in terms of fraud, what's keeping you the most busy? The greatest challenge, I think, the financial services organizations face across all lockers, consumers become completely intolerant of any friction in those interactions that they have with financial services companies. And if you get it wrong, you know, that can have enormous impacts. And you know, you get it wrong from a fraud assessment to your point, there's an open checkbook and potentially for that individual, they move into your organization. Within credit info, we've got such a diverse range of markets, economies that are just starting out in terms of digitization, where, you know, the governments have recognized that digitization is the way to stimulate economic growth. So you have the banks moving towards that and other lending companies moving towards having those ditches on boarding processes. If you can establish that on boarding process, which is nice and clean, friction free, you're getting great PR on your app reviews, people are telling their friends that it was really easy to open an account. It just took me five minutes and enough to go into a branch, say here in the UK and that's where I'm in the minute. You know, I suppose we've been telling that story now and we'll well advance in terms of the digitization of our economy. Nobody would consider going into a branch nowadays to open a bank account or to apply for a credit card. Where is the first port call or, you know, organizations like Revolut at based banks, we just completely intolerant of that friction. But if you look at it, then across other markets, you have places like Kenya, even to get a mobile phone number, you need to actually physically go into an agent's office with your ID card and then your phone number is linked to your identity number. So you know, the level of friction that Kenyans are used to is far greater than we would ever stand up for in the UK. But it also comes with vulnerabilities and, you know, the fact that you can pop down to downtown Nairobi and pick yourself up a very genuine looking ID card. You know, fraudsters find a way around those controls. So, you know, whilst you've got those physical controls in place, you know, there are vulnerabilities in the system with the over-reliance on the ID card. Yeah, fraudsters will find a way. So you are always having to change strategies and keep on top of it, I think far more frequently than other areas of the bank, you know, scorecard building and scorecard building for the most spot. So something to keep you busy. I mentioned a today, you know, we skipped quite quickly at the front. The apparition of slow down a little with talking about credit info. We have had Paul Burak and your Katarine on the shows of people. I'll probably somewhat familiar, but credit info for me in terms of the big credit information companies around the world is unique in its mix of countries that it operates in from conservative, wealthy, quite structured Scandinavian countries to some of the more emerging, newer emerging African economies as well. So you've got to deal with not just fraudsters that change every time you close one loophole. You've got to change with whole, you know. market cultural differences, stages of development, acceptance of friction. So you've just released a global fraud and unentity program in Kenya. They're talking a little bit about that, but also talking about how you think about your strategies globally and how you build a product across such diverse markets and make sure you catch and fraudsters in all of them. I mean, I guess speaking to your point about geographical diversity, there's probably three components to that diversity that we'd say into a county. You're point about, yeah, we've got mature Scandinavian, Baltic, digital economies with strong national identity digital identity schemes versus emerging digital economies that are starting that journey to digitisation. And then you've got a robust AML regimes across Europe with rigorous enforcement actions and fines of tens of billions of euros for breaking the AML rules versus markets where they are on the gray list from a money laundering perspective because they haven't got adequate controls in their financial services sector. And that's not always about customer due diligence, but customer due diligence and the kinds of solutions that we're taking to market. Certainly speak to that gap that has been identified those strategic deficiencies. And what we see is that fraudsters have probably targeted European markets more because of that digital maturity and we'll see more of that within these emerging digital economies as they grow. But at the other end of the spectrum, we've got an almost overreliance on some of those digital identities across the European zone. And I think that breeds complacency because the organisations think those identities are watertight. And our solutions really that we've built out speak to both ends of aspect from in terms of maturity. We've got pressure for real time decisions and intolerance for friction. And we've got to make sure establishing a trust for a consumer in our clients organisations because all we're really trying to do is answer three simple questions. Is this presented identity a genuine identity? Are my interacting with the owner of our identity? And are there any risk factors associated with the identity out being presented with? That's all we're trying to do. We're trying to make life difficult for bad actors by establishing a hostile environment for them to operate in by introducing a significant number of controls. We're trying to take friction away from genuine interactions by basing that upfront decision making on inside that we derive from data, trying to make life easy for forward fighters is much that we're exposing a whole range of different services that are applicable. One of the things that attracted me about role credit info was the agility and flexibility in their existing technology stack. It's quite easy to add new services to our solutions. So if you've got a range of local sources of insight within a particular geography, they can easily be added alongside alongside the data that we've got from global services. And also we've got a standard credit bureau data structure across the globe. You can drive inside our data, bringing as many data sources together that are applicable on a global basis based on credit info data that we hold and other data sources. And then using ML models and other data science techniques then to drive predictive insight, establish you in that hostile environment for for instance, there's another bad actor's. I mean, I'm including Mollilandras within this as well. We're just putting so many controls in there. It's next life difficult for them, it's so difficult to synthesize a genuine identity across a whole range of different basis sources. So we're looking at not only credit bureau data, but that combined level with local services. And then also looking more broadly across alternative data sets. Identity today is far more than your name and debt of birth and address the national ID number. It's any linked email addresses, any linked mobile numbers, any linked messaging services or social media services. Have you got an Amazon account? Have you got an office 365 account? Is the data consistent between all of these different sources? And that's the solution that we're bringing together using the flexible technology to assemble that staff controls for organizations to deploy aligned with that particular risk profile? Yeah, it's funny to hear how sophisticated it has become. I remember back again, sort of sound of the very old here, but back in my day, I was trying to test something out for one of our very early fraud detection models. And I found out at the time we were running national elections and there was a website online where you could put your ID number in and it would tell you where you were registered to vote and it showed you your home address linked to that. So I could just generate made up, but possible, probable identity numbers. If they existed, it told me the name of the person there address and I could then have gone on and started using that. But now I'm hearing you've kind of got to play around with time and operate across all these systems. But of course, fraud's gone a lot more sophisticated since the people are able to do that. And it's kind of eye-opening. I mean, one of the key trends that we're seeing and this is probably an established trend in the US, but it's not constrained now to mature economies really. Is this rise in synthetic identity fraud where you've got fraudsters building an identity up from scratch, creating an email address, linking it to a phone number, how does a fraudster create a hundred email addresses last week and then makes them look like they've genuinely been existing over time and life over time. How do they synthesise the presence of that email address in a data breach five years ago alongside the phone number that they've already provided. And this is what I'm saying about establishing that hostile environment for fraudsters. Fraudsters by nature are lazy, right? Otherwise, they get proper job like the rest of us. So you make their life difficult, they'll go and find an alternative target. But yeah, it's not a game that's won, you know, fraudsters innovate at pace, outside any regulatory constraints. You see them harnessing new technologies, looking at using, for instance, AI to generate fake identities or to export organisations on board in processes at scale to try and identify vulnerabilities. You know, we've seen the convergence of cybercrime fraud, money laundering, organised crime coming together as kind of fraudsters of service offerings. You can go out and procure yourself a fraud service in the same way as a genuine organisation can go out and get a Microsoft Assault Clired Instance. There's always going to be this constant fighting battle where we're going to need to be addressing new emerging threats. And applying AI in our only case system is just as important as you know, fraudsters applying it in their activities. Yeah, so applying machine learning, looking at big data analysis and leveraging data science techniques using natural language processing to sort of like better understand some of the interactions that you're having using AI to simulate attacks on your organisation. It's always going to be this kind of mouse game. That's one of the, one of the things that we're considering as we're building our solutions is that we have to retain that level of flexibility. Yeah, and you spoke about the convergence there. I think it's also evident in our view as laypeople if we just look in the things like newspaper headlines. The UK government says, you know, talking about fraud and fraud prevention strategies. We see a lot in the press about cases where customers have fallen victim to some sort of scam, have lost some money, whose responsible how much should the banks be covering. And yeah, government's announcing put steps in programs to address fraud. Perhaps because of that coming together of all types of fraud, it's no longer an isolated credit card issue or KYC issue. It's all being done together by the same sort of people. But do you see around the world that governments are responding and kind of putting in place efforts to shore up their ID systems to do whatever they can to make the ecosystem more protected or how you're seeing those words translate into actions. So the UK government have sponsored a program to develop a governmental fraud and cybercrime strategy and that speaks to that convergence. You know, they're seeing that coming hand in hand. Now, you know, the controls in market are less mature and perhaps than they need to be. But we're definitely seeing, we're definitely seeing government programs across a number of markets, Sri Lanka's another market. and the government, they're looking at introducing a digital identity ecosystem in the same way that we have across Europe in the form of the Idaas. Those ecosystems come with some significant barriers to overcome interoperability and adoption across industry. The UK population don't like that kind of ID scheme, but you can see a well clear of it a few attempts in the past. And then he takes the liability if there's fraud based on an identity that's been created elsewhere. So I think they come with some challenges to implement an over-align some of them we see in some of our Baltic countries that the smart ID that's used there. People just see something pop up on their phone as an alert and they just go, "Oh god it's that again right and click it and all of a sudden you you know you're into that fraud space." So this challenge is there and I think you know as though those ecosystems do emerge in some of these markets, then we're still looking at this same set of controls that we need to leverage yield to establish those identities in the first place. So rather than providing these digital footprint assessments and credit bureau scores based on historical data to clients, these are serious contributions into those digital identity creation mechanisms. Again, it's just to make that hostile environment more complex. You make it the harder for the fraudsters, but also it could remind of how some of the friction in getting this right comes from cultural things. You've got no control over that are just how countries developed historically. So South Africa, we always had IDs. Probably I'm guessing I don't know their full history but likely related to a part aid. And so probably put in place for bad reasons to keep track of who's in which group. But it was just normal for us. We had ID books. We still have ID books and you've got a national ID database. And it's quite convenient. Privacy concerns aside, your countries like the UK, they've resisted every sort of move towards an ID card. Your countries like India were there sort of a very complete one system with all the biometric stored in place. When I was living in Denmark, I don't know if it's the same now. My ID card didn't have a picture on it was just my name and address. Likewise in Hong Kong, my driver's license has no picture on it. So you've got all these weird different systems that are one extreme or another somewhere in between that have developed. And people because it's around your identity, you hold it very tightly to what they feel is right culturally, which only complicates. I'm sure there's sort of work that you do when you roll it out. I think you summed it up in a nutshell there in terms of what credit and foes approach is to these challenges that organizations face in terms of lending money to strangers. Love the name of the show by the way. Rob credit and info is around the world. If people want to look at some of these fraud protection tools that you've got available, they want to maybe read up a little bit on their new tools you're rolling out in Kenya. Where do they go to find out more to maybe start a conversation? So in terms of more detailed Kenya specific product information, you know, Kenya website. And there's got the detail on there. We've gone through some changes that credit info over the past couple of years and we've established a group level CEO and a team sitting underneath us, which is where I sit. So we've now got a group team looking after the businesses and we're matrix then into all the different markets that we operate in. So we've got a group website, createinfo.com. That's got links to all of the individual local sites. We've got group LinkedIn and local business LinkedIn. We're quite active with a minute around providing insights into what we're doing in Kenya as we move out to other markets. You know, you can be next and then on to the restafrican markets Sri Lanka, cross central Eastern Europe over the coming months. We'll see that the local website is gal dated. We've gone the group website. We've got a credit info chronicle. You were posting information about products on there. So our group website group LinkedIn account and then the local the local website. So I'm quite happy for people to pick up directly with me. If they want a conversation about some of the solutions or some of the challenges that they're facing in markets that Wow, we've not got to yet. Yeah, awesome. So I'll put your your links in the show notes as well. I'm on the the chronicle page now. There's a really nice article explaining the new solution that you've just rolled out in Kenya there. Discussing it's going to mix of data sources and technology sources that are involved. So for anyone who wants to kind of have that more slower read through and get their head around what that's offering they can do that. And I think what I think what people will see is they you know, it's like Staggs boring what credit info to is that the fraud and identity initiative I'm responsible for is just one over a number of initiatives that we're launching at the minute and the fraud and identity as solutions sit quite nicely alongside our business information solution. So an organization wanting to do due diligence around beneficial ownership or the officers within an organization can use all business information tool alongside that the fraud and identity solutions and then consumer facing solutions that bring credit information together with the identity information and I and anti forward education. So you know, there's a whole range of initiatives that we're we're launching what I've launched. Really glad you brought that up. I think it's the benefit of all the data coming together and being analyzed with machine learning techniques and with sort of very careful eyes to look for these patterns. You know, that data can tell us many different stories. So we're talking today about fraud, but if somebody is wanting to wonder how do they better use credit bureau information in one of the countries you work in to better lending decisions to meet their compliance needs better or more efficiently. The sort of solutions is that certainly the skill set and data set that we're talking about today for fraud could equally help in other areas of your lending business. So we'll put the links to the credit info group. I mean, if it's not fraud, you're interested in, but you want to learn how something similar could help you with your school card, but they could have the cash may be it's certainly worth looking at because it's this understanding of big data that describes our lives. So Rob, before I let you go, I know that the team's always kind of traveling around and talking to people face to face as well, but is there anything in particular on the horizon that you you want to draw people's attention to that they should keep an eye out for any kind of new things happening in this space? I mean, I think I'll speak specifically for the Ford initiative from the minute. You know, I would probably just dip in your toe in the water. So you know, my focus over the coming months is really to build on that foundation and then be adding additional face sets to these solutions, building our capability around device around biometrics. You know, there's a need for that in many of the markets that we operated. So yeah, that's that's really going to be my focus over the coming coming months is to make sure that we've got that full stack service, but also know looking for opportunities to leverage these kinds of solutions beyond lending to strangers. How do you identify bad actors that are already within your organisation? So moving out of that customer management journey. So those are sort of like the big ticket items on the agenda for a minute. Yeah, so really exciting times, actually, as you think about what's possible. And so yeah, yeah, anyone listening who's got a slightly different fraud problem, feel free to reach out as well and see if there's something coming up or there can be done to solve that, I think a lot that will keep me interested in following these stories. Great markets to hear that innovation in Kenya and Uganda to really exciting markets, Sri Lanka's doing cool things. So I encourage everybody, particularly obviously, those working in fraud prevention to have a look, but even those who aren't who maybe aren't familiar with how high tech innovation is in these markets that often aren't talked about. It's worth seeing what sort of solutions have been rolled out, now in places like East Africa, Central Eastern Europe, Southeast Asia. You know, this is not the old world where we slowly hand down systems and expect when I was working for a banking Kenya, we had a special old computer that had to be protected with our lives. I think it was an old 486 that was running and it was protected because there's only one old enough to run the software that ran the credit card system. The system may have been built in the 80s. Every year, some guys came out from Spain and they tweaked the coding, but it was, even in those days, would have been a 30-year-old piece of software running the bank. Now we're talking about machine learning, our tools been rolled out in the same market. So lots of cool innovation happening and lots of inspiration. So Rob, thank you so much for coming on. Always love to hear what credit in first doing. Yeah, brilliant. Thanks to our ambient branders, nice to catch up again after all these years. And thank you all for listening. Please do look for and follow the show on your favorite podcast platform and share the updates widely on LinkedIn, where lending notes are found in our largest concentration. Plus send me a connection request while you're there. This show is written and recorded by myself, Brendan LeGrange in Brighton, England. Show music is by I.M. Wake and you can find show notes and written transcripts at www.artilandmoney2strangers.show. Or just www.htlmts.show and I'll see you again next Thursday. Music.