Experts from Featurespace, TSYS a Global Payments Company, Mastercard, and Worldpay from FIS discuss the role of data and self-learning in fraud and AML products

In part four of this blog series, our experts analyzed the current and future trends in fraud and fincrime.

  • Dave Excell, Founder at Featurespace
  • Dondi Black, SVP, Chief Product Officer at TSYS
  • Liam Cooney, Vice President at Mastercard
  • Mina Khattak, Director, Crypto & Emerging Business, Worldpay from FIS

In part five, these payments powerhouses emphasise the crucial role of machine learning in tackling emerging fraud and financial crime risk.

Pragmatic, measurable, explainable machine learning

Excell: “When we look at the machine learning deployed by the fraudsters to do things like deep fakes to impersonate identity, we definitely need to have better tools to be able to find out what they’re doing. To make the transaction safer for everyone.”

Cooney: “I think that look we want to make sure that we’re providing safe, simple, transactions and enabling trust across all the interactions that take place in the payment ecosystems. I think though some machine learning I feel like it’s table stakes at this point, because of the growth in volume of digital payments, the number of transactions, the amount of data that is there, you cannot put any good layered strategy that’s going to be effective. You will be a weak link and you will get exploited. You’ll pay for it and your customers will pay for it, and so I, I think that it is table stakes, but it needs to be done in in a way that is pragmatic and drives measurable results and is explainable.

I think that we’ve found success in in doing that by trying to boil it down to where we have visibility of certain transactions and can be a connected intelligence layer as transactions and accounts move across an ecosystem.

One of the ways that we’re starting to think about that now is: instead of developing solutions that are just focused on one particular payment type or rail, start thinking about how do these behaviours and accounts transact across multiple transaction types? The reality is a fraudster does not care how you make the payment; they just want the money. They’re targeting the payments on a card, they’re targeting it through a push payment, or they’re pushing it through crypto. These are all the three kinds of main pillars of payment types that we’re focused on right now and trying to identify what’s really successful? In terms of ATO on cards or cyber risks as it relates to our entire card network…and we start to bring those proven technologies over into the account space or into the crypto space which is even more nascent.

There is a lot of work to be done, but it’s very promising that we’ve had that success in some of these other channels because we know it can be replicated. Somewhat frustratingly, the technology and the capabilities are potentially keeping us from realizing this as an industry. It is often the lack of basically governance (model, regulation, and rules) in which participants can share the data that they need to do this. I’m hopeful, but I know it’s going to take time as each market will go through its own process to allow this to happen.”

Power in numbers

Excell: “If you think about the number of different types of customers that we serve – from very large multinational financial institutions with lots of scale and lots of sophistication to smaller organisations looking at new payment flows and starting out – you see the different requirements for fraud and financial crime prevention.”

Khattak: “It’s going to be very difficult for an individual merchant, an enterprise level merchant, or a Small to Medium Business to fight fraud independently. You can leverage solutions that have broader data sets: merchant acquirers have the resources, and we have the incentive and the data to be able to do this effectively. The extent to which you can embed a solution on your payment platform, I think you’ll see real impact and a reduction in fraud. Rely on those that can do this effectively before trying to do it on your own.”

Cooney: “I think that some larger, more sophisticated institutions their demand is really more for the intelligence. It needs to be an API. They’re more interested in the metadata to augment existing models that they have and see if they can drive better results. On the other end of the spectrum, you’ve got smaller institutions that are starting from there zero and they need end-to-end enterprise fraud or AML solutions, and that intelligence needs to come over time. I think that the reality is we need to be able to provide those capabilities to all of our customers. We tend to focus more on the consortium, network intelligence and we work with partners like Featurespace on bridging any of the gaps that we have in terms of capabilities that we can offer to our customers to make sure that we’re putting the best capabilities in our customers hands and meeting their needs.

Global expertise

Black: “Our goal is to ensure that no one is starting from Ground Zero, that we’re able to bring the depth and the breadth of our global expertise, the line of sight that we have and bring that to bear. The reality is today we are all we are all operating in a global economy, so this notion of how we service our clients in a cross-border way, with no friction, and protect cross-border experiences is increasingly important.

In this space it’s going to continue to change and evolve rapidly, our remit is to be that connective tissue, to be that enabler of the entire ecosystem and ensure that no matter what the form factor is, what the channel is that our clients, be they financial institutions, large and small fintechs, corporates, SMBs, that they are able to accept, receive, initiate the transfer of value in a way that is simple, seamless, and secure.

Excell: “Fraud has continued to grow. Unfortunately, we’ve all done a lot of huge amount of work, but there’s still a lot that we can continue to do. The fraudsters have access to data and technology, and consumers continue to want a seamless experience. We’ve had some great partnerships to be able to move drive improvements and I look forward to continuing to do so.

The Business Case

It’s clear that fraud prevention and AML is no longer an afterthought or siloed from the payments services themselves. And for good reason. Not only does great fraud and financial crime analytics allow issuers, acquirers, and networks to optimize their payment services: best-in-class detection enables them to productize their investments and package them into value-added services.

Every payments player in the ecosystem looks to be first to market with new payment types, but we’ve learnt the lessons of the past: it’s not ethical to launch new consumer services without strong fraud and financial crime protections, and it’s just not good business sense!

Learn more about Featurespace’s white label solutions for payment issuers, processors, and acquirers.


Part 1: The Next Big Thing in Payments as a Service? Fraud and AML

Part 2: Product Design for Fraud Prevention and AML in Payments as a Service

Part 3: Choosing the Right Technology for Fraud Prevention and AML in Payments as a Service

Part 4: Current and Future Fraud and Money Laundering Trends