Q&A: For Effective Financial Crime Prevention, Build a Better Mix of Machines and Humans

Bill Millar (CCI): What are the challenges and opportunities in financial crime compliance?

Leslie Bailey (LB): For many businesses and financial institutions, financial crime compliance is top of mind. Covid-19 introduced complexity and risk into supply chains. But with the Russian invasion of the Ukraine, that complexity is further amplified. So, if you look just within the past month and a half or so, you see sanctions becoming a critical issue, with companies scrambling to keep up with who they’re doing business with.

In addition, financial crime risks are rising around the edges of business. Think about the end of cash, and how people view alternative payment methods like cryptocurrencies — particularly considering there are entire countries looking at alternative currencies. These are issues that have been around for quite a while, but they’re today gaining a lot more traction.

Also, think about the expanding focus on anti-money laundering (AML) and issues in unidentified ultimate beneficial ownership (UBO) that are all part and parcel of financial crime avoidance and compliance.

CCI: So, what’s the insight you wish to share on these fronts?

LB: I want to level-set on the best-use technology to manage corporate crime. When people think of advanced technology, they automatically go to artificial intelligence (AI) and machine learning (ML). And as world behavior began changing with Covid, these were two of the technologies companies that have been top of mind for tasks, such as more appropriately identifying their risks.

But inherently, there’s risk associated with automating anything. Right now, I know that a lot of people are leaning toward machine learning. But the issues with that: What happens, and how do you pivot when behavior changes?

CCI: What’s your alternative suggestion?

LB: In our view, we say advanced analytics is likely the more useful path to take. Advanced analytics enables you to identify who you’re doing business with, or who you have in your book. Where we really see the value is making sure on the front end, you have the appropriate filters. You must have the appropriate recognition of an individual as a bad actor — and that’s advanced analytics.

If you want to achieve zero tolerance for financial crime, what you want to do is make sure that you have humans who are conducting the appropriate investigations. You don’t want to miss anything coming through the front door because you’ve automated too many processes. But you also want to avoid producing false alerts that have the downstream impact of triggering a lot of unnecessary due diligence. You need the right balance of human and machine.

CCI: How should businesses achieve this balance? 

LB: You want human intuition at the front end. But beyond that, I think organizations must look more expansively across their businesses into how they collaborate internally. They need to be able to bring all their information together and combine what they learn from an investigative process to really see their exposure. And they need to ask questions about their own data: Does it indicate this activity could be potential money laundering? Does this other data indicate potential human trafficking?

They also can use advanced analytics to identify irregularities: Why is the person logging in from Iran or Ukraine when this is an account that was opened by a college student in New York City? When people leverage data they already have, along with risk-appropriate external data, they can run smarter.

CCI: Do you see any other changes, risks or opportunities on the horizon? 

LB: A big change coming, and very important in this area, is the adoption of ISO 20022. This is a new global payment standard [within SWIFT that is intended to provide participants with a rich source of information in a format that’s easily digestible. So potentially you’ll have fewer failed payments resulting from better insight into customers.]

For example, what if you have a name that is similar to someone else’s who has been sanctioned or is on some other list? Since we have more information, when you want to open an account at a bank or move money, we’ll have enough information to properly identify who you are, and this will help us prevent people from [getting caught up in the system] when they shouldn’t.

CCI: How soon will this be up and running?

LB: Adoption is to begin later this year. The dates have been moved a couple of times, but it is now slated to begin in November 2022. Presumably, organizations are in a ready mode, as they should be. But they should be communicating with their customers about what’s coming: Financial institutions can run more efficiently and create innovative new services. Customers will reap the benefit of better service and experiences. And firms will experience fewer false positives, so they won’t be wasting resources looking into innocent transactions.

Of course, it may take a year or two to get everyone on board and to work out the bugs. But this will be an important set of improvements in international payments and will help in the fight to detect and prevent financial crime.

CCI: Any final words?

LB: Just that fully automated processes are built around what people have seen, not around what hasn’t happened yet. But behaviors are always changing — and today, rapidly. So, automate, but focus on advanced analytics. You absolutely do not want to overlook the vital importance of human intuition.

This interview has been edited for length and clarity.

Leslie Bailey is the vice president of financial crime compliance strategy at LexisNexis Risk Solutions, a provider of technology and analytics solutions that helps their customers assess and predict risk. Based in Alabama, Leslie directs marketing strategy for the LexisNexis Risk Solutions’ financial crime compliance solution in the United States and Canadian regions. A graduate of Samford University, Leslie holds many professional certifications, including Certified Fraud Examiner (CFE) and Lean Six Sigma Green Belt.