Authorized push payment scams – where victims are tricked into sending money to criminals under false pretenses but of their own accord – are a growing problem. Banks are testing biometric technology to help flag unusual customer behavior.

A new report from the Global Anti-Scam Alliance found that scams increased by 10.2% globally in the past year, reaching 293 million reports and leading to a record amount of money lost by victims. (The study also notes that this underestimates the total problem, because scam reporting remains low.)  

Much of this fraud involves authorized push payments (APPs) – where victims get swindled into willingly sending money – which are notoriously difficult for banks to deter. But as regulators in the US ramp up pressure to add protections and grant reimbursement to victims, there’s a growing push to use biometric technology to slow down these scams.  

While biometrics in banking usually involves face or fingerprint scans, several tech firms are deploying machine learning to spot customer behavior that could be a sign of potential coercion. Suspicious traits like hesitation or choppy typing can prompt banks to inject additional warnings to customers or even delay or prevent payment. 

“Behavioral clues that are out of the norm – such as fumbling and hesitating while entering unusual amounts to be sent to unfamiliar recipients’ phones or accounts – are some of the red flags biometrics spotlight,” an exec at tech firm Callsign told American Banker. Several banks in the US, Canada, and the UK are testing Callsign’s technology, while a similar firm called BioCatch has had an uptick in interest too. 

Customer education is currently the main tool banks use to mitigate APP fraud and biometric intervention offers a promising additional option.