Can AI combat complex cross-border payment crime? | PaymentsSource

Banks add new Artificial intelligence to combat fraud These include payments, email phishing, identity verification, document analysis and other threats. However, using AI to protect cross-border payments brings other challenges.

AI is a great tool for curbing financial crime, especially in domestic consumer payments due to the repetitive nature of consumers' transactions, said Stephen Grainger, head of data and analytics and FCC at the Society for Worldwide Interbank Financial Telecommunication (Swift). Swift provides the primary messaging network through which international payments are initiated.

“A person tends to make the same types of recurring payments,” Grainger said. “It would be easy for the issuing bank of your credit card or your debit card or for Venmo to see [recurring payments]. You can start modeling this behavior.”

However, cross-border payments are largely corporate and lack the regular cadence that prevails in domestic payments, which makes these transactions more difficult to modelAnd the amount of corporate payments is usually higher than domestic payments, which also means higher potential losses.

“This is the complexity dynamic that we need to get a handle on,” Grainger said. “How can we rethink the way we think about fraud and anomaly detection?”

In addition, much of cross-border payment fraud is operational fraud, Grainger said. For example, a fraudster might impersonate a CEO and ask him to transfer money to a new customer.

“This also makes it more difficult to detect cross-border fraud because it is most likely an internal operation,” he said.

Swift plans to use AI in “multiple ways,” Grainger said. The company uses machine learning to allow companies to embed rules into workflows. For example, a company could set a rule that says, “When I pay someone new, hold the payment.”

The company, based in La Hupe, Belgium, is also testing federated artificial intelligence and machine learning, which uses modeling instead of data, in the hope of reaching banks that are reluctant to share confidential information.

“You take the model [and] Train it on a dataset and instead of transferring the data, transfer the model,” Grainger said.

Swift is currently working with “some banks” to define what this model could look like, Grainger said. “We believe that [federated learning] The model is gaining traction with more regulators as they begin to think about how to solve the challenges and frictions associated with cross-border payments,” he said.

Finding ways to exchange information between the parties involved in the cross-border transaction is of utmost importance for Using AI effectively to combat fraud, said Ben Turner, president and CEO of payments fintech company Verituity, which works with Mastercard and BNY Mellon, among others.

“The way we [AI] is to identify, establish and test the reliability of relationships throughout the transaction,” Turner said.

“Whether it is the relationship between buyer and supplier, the relationship between the supplier’s administration and the supplier, [or] When you look at the relationship between historical payment activity and today's, you try to identify anomalies,” he said.

Sharing information between payment parties could help counter the rise of deepfake fraud by leveraging data known only to those two parties, similar to the out-of-wallet questions financial institutions use to combat consumer fraud, Turner explained.

AI is also often used to verify the direct recipient of a payment and their acquaintances. This is called “entity resolution,” says John Meyer, Managing Director at Cornerstone Advisors.

“Today, there's more sophisticated AI that says, 'The person you're sending money to isn't on any of these well-known watch lists… but they're really good friends with them,'” Meyer said.

But fraudsters are also using artificial intelligence and other new forms of machine learning to drive deep fake fraud. For example, fraudsters can record a person's voice sample and use it to ask an employee to transfer a payment to a new customer, or to retrieve important information such as passwords by using known information about the person, Meyer said.

In cross-border payments, which often involve numerous financial institutions in multiple countries and jurisdictions, the risk of fraud increases with each additional party, says Luke Penca, managing director of the consulting firm Capco.

This makes data communication between the parties all the more important, said Penca.

“We've had a rules-based system for a long time, but I think AI is really helping banks make sure they have a really good, vigilant process there. The bad guys have learned and are equipped with AI,” Penca said.