Behavioral Analytics Inject ‘Smart Friction’ Into Battle Against Scams and Payments Fraud

Scams against banks and their customers have exploded.

Fraudsters pull at human nature in a bid to gain access to accounts and drain them — pleading through texts, phone calls and artificial intelligence (AI) prompts for donations, romance, bail to get out of jail and more.

Scammers are industrious, becoming more businesslike, moving beyond blast emails toward a personalized approach as they pick their victims.

The PYMNTS Intelligence report “The Impact of Financial Scams on Consumers’ Finances and Banking Habits” found that two types of scams inflict more financial damage than average. Investment scams have a median loss of $1,104, and romance scams have a median loss of $1,996. Romance scams also string targets along for an average of 3.6 transactions, which is nearly twice as many as other methods.

Featurespace founder David Excell and Gasan Awad, senior vice president of enterprise fraud product management at PNC, said successful defenses depend on pattern recognition, expanding the channels through which banks communicate with one another to keep abreast of emerging attack vectors, and educating customers. The right data, rendered in real time and scrutinized with behavioral analytics, can pull a payment right when the request is being made, protecting all stakeholders.

“[Scammers] have an ROI in mind — and they’re doing segment testing, trying to find out what’s most effective, where the weak links are in the institutions, and where they can attack,” Awad said. “There’s an evolution, and we need to be more on the offensive than the defensive that we have been.”

Banks strive to give consumers the best info needed to make the best payment decision, and they must navigate the balancing act that comes with safety and convenience for an optimal end user experience, Excell said.

AI is a weapon, and fraudsters are “using it for deepfakes, impersonations and automated attacks … it’s not just email … and it’s coming through social media apps” among other channels, Awad said. It’s a trillion-dollar issue, where as much as half of the adult population encounters a scam on at least a weekly basis.

Featurespace has been using AI and machine learning to understand consumers, how they pay, and to identify “behavior which is suspicious and uncharacteristic of the ways in which they would interact with a bank or credit union to make payments,” Excell said.

That data is relayed back to banks to help them decide when they should introduce friction into the process. It challenges the customer in a way that generates trust that the bank is monitoring their transactions, especially if they are making a transaction for the first time or making a large-dollar payment, he said.

As banks and providers build frameworks and mechanisms to step up fraud defenses, “customers want to transact in manners that are safe and convenient, but that does not mean ‘totally frictionless,’” Awad said.

Questions regarding whether customers want to proceed with transactions and whether they truly “know” recipients can keep individuals safe from an “urgency trap” that scammers love to set, pressuring would-be victims to hurry up and send money or account details, he said.

The same data that informs banks about customer-level behaviors can point to where scammers are hiding — identifying activity as money moves into their accounts or to certain IP addresses, he said. Social media and telecommunications channels can provide hints as to what’s happening behind the scenes.

“The ability to look across your portfolio and make sure you connect the dots is critical not only in preventing [fraud] but also in knocking out bad folks,” Awad said.

The consortium approach to fighting fraud — where banks keep one another informed in real time — remains among the best ways to capture and short-circuit the monetization of scams, Awad and Excell said. There’s a place for regulators and law enforcement to help define how, when and what information can be shared.

“The more we work together, the better,” Awad said.

Sharing information blunts “the asset that the scammers have … the ability to be able to capture the money, move it, and then sort of take it back out of the financial system for their own control,” Excell said. “The harder that we can make their job, and make their ROI smaller and smaller, we’ll start to push [fraud] somewhere else.”

The post Behavioral Analytics Inject ‘Smart Friction’ Into Battle Against Scams and Payments Fraud appeared first on PYMNTS.com.