Feedzai Expands Digital Trust to Help Banks Detect AI Agents at the Front Door via Cryptographic Device Binding

Fraud-detection vendor Feedzai has expanded its Digital Trust solution to help banks identify whether a banking session is run by a human or an AI agent, including agents operating with valid customer credentials. Announced on 12 November, the capability combines persistent device fingerprinting, a cryptographic device binding that forms a trust triangle between customer, device and bank, and fraud scores computed from a cross-bank network view and shared intelligence. The company frames the posture as blocking malicious agents while trusting those chosen by the customer, with consent control in the bank's hands, and positions the same signals as enabling stronger authentication with reduced reliance on SMS one-time passwords.

The launch lands two days after BioCatch introduced AiRE, which approaches the same problem through behavioral analysis. Two major vendors entering the same category within 72 hours suggests the market treats agent detection as a coming requirement rather than a niche experiment. A successful login no longer implies a trustworthy session.

All capability claims are vendor-reported, with no reference customers named and no general-availability dates for individual components given. For banks in markets where OTP SMS remains the dominant second factor, the combination of agent detection and device-bound authentication is likely to arrive through global vendors before local policy catches up.