Banks struggle to fight fraud with AI, reusable ID could help

The vast majority of financial institutions – around 90 percent – are leaning on some form of AI to combat fraud and financial crime. Among them, 30 percent are using AI for identity verification, according to a new report from financial crime prevention and behavioral biometrics company Feedzai.

The report, titled 2025 AI Trends in Fraud and Financial Crime Prevention, delved into the ways banks and other institutions are combating fraud, including threats posed by generative AI. The insights are based on a survey of more than 560 anti-fraud professionals working in the financial industry.

While the survey shows a strong adoption of the technology in the financial industry, AI represents a diverse set of tools that can be used both for good and bad. The rise of generative AI, for instance, is fueling the rise of fraud: 60 percent of respondents said that fraudsters are using the technology for voice cloning scams, says the report.

In addition to digital identity verification, half of financial institutions are applying AI-powered solutions for scam detection, while 39 percent of them are using it to zoom in on transaction fraud. Nearly a third are using AI for automating Anti-Money Laundering (AML) processes and customer banking journeys.

AI is also proving effective for tackling false positives, with roughly one-third of respondents witnessing a large reduction of false positives of 40 to 60 percent.

Challenges remain, including data privacy and consent, reducing bias and justifying costs as implementing and maintaining AI systems can be expensive. Explainable models and interpretability tools are essential to meet compliance and build trust, the report adds.

Identity verification is getting democratized: Sumsub

The financial industry has been reimagining digital identity and making finance a seamless part of people’s lives. A part of this process is the democratization of identity verification, which represents a shift from bank-controlled models of the past, speakers said at the Money 20/20 Asia conference held in Bangkok, Thailand, in April.

Aside from representatives from the financial industry, the conference invited identity verification firms such as Sumsub.

Traditional banking identity models meant that each institution maintains its own Know Your Customer (KYC) silo, forcing customers to go through unnecessary verification processes. Today, however, customers can use a secure ID that helps them get onboarded and access banking services faster, all while maintaining compliance with regulators, according to trade publication CDOTrends.

Sumsub’s self-sovereign identity (SSI) system, for instance, allows customers to verify once and use their identity across different services, all while maintaining control over data, says Sumsub’s Vice President for APAC, Penny Chai.

“The controller is me, you, and everyone else,” says Chai.

Banks are warming up to this type of digital identity, not just because of convenience but also because of fraud. According to Sift’s Q1 2025 Digital Trust Index, sectors like fintech and online marketplaces are seeing high rates of payment fraud attacks, while the total cost of digital commerce fraud could hit US$107 billion by 2029.

Generative AI has been fueling the rise of synthetic identities. Companies such as Sumsub are using AI to detect them through multilayered security approaches, including device fingerprinting and behavioral analysis, adds Chai.

In a recent webinar hosted by Biometric Update, Mitek and Goode Intelligence presented on increasingly popular attack types in biometric payments, including bust-out fraud and account takeovers (ATOs). The webinar is available to watch on demand.

Article Topics

banking  |  digital ID  |  Feedzai  |  financial services  |  fraud prevention  |  identity verification  |  KYC  |  Mitek  |  reusable digital ID  |  self-sovereign identity  |  Sumsub