Feedzai, the AI-native financial crime prevention specialist, has broadened its Digital Trust offering so banks can identify when AI agents, rather than people, are entering customer accounts.
The upgrade lays the groundwork for agent detection at scale through three core elements: more durable device persistence, a fraud score drawn from Feedzai’s network-wide view and pooled intelligence, and cryptographic device binding that links the customer, their device and their bank in a single chain of trust.
Each element also works independently, allowing banks to reduce needless step-up checks, lower authentication spending and better defend against attempts to bypass multifactor authentication (MFA).
Feedzai’s argument is that a successful sign-in no longer proves a session is safe. AI agents can log in using legitimate credentials, while fraudsters use hijacked devices, account recovery routes and social engineering to manipulate genuine customers.
Digital Trust addresses this by assessing endpoint, device, behavioural and network data both before login and across the entire customer session.
Identifying an agent depends on recognising the device, relying on trustworthy signals and understanding the purpose of each session. Persistent device fingerprinting and network data allow the platform to spot repeat activity over time, and the risk score taps Feedzai’s broader network to reveal patterns an individual bank might overlook.
Cryptographic binding then connects each session to a known customer and device, showing banks whom an agent represents.
The system flags agents by pairing the information they disclose about themselves with fingerprinting and behavioural indicators of automation. Once an agent is identified, banks can recognise repeat user-agent pairings and skip further challenges where policy permits, while retaining control over which agent activity is permitted and how consent is managed. Feedzai plans to extend its agent detection tools over the coming months.
The release also strengthens authentication for human customers while cutting dependence on SMS one-time passcodes (OTPs). Because device binding connects the customer, device and bank, a trusted device can serve as a possession factor.
Digital Trust additionally guards passkeys against bypass by verifying that a registered device is present and inspecting it for compromise or manipulation. This covers login, account recovery, new device registration and call centre verification tied to a digital session.
According to Feedzai, using shared signals uncovered 44% more fraud among the riskiest 0.1% of sessions without raising false positives. The firm also reported 6,000 fewer unnecessary authentication challenges per million logins compared with rival technology, and a 41% drop in SMS OTPs.
Feedzai chief product officer Pedro Barata said, “A leading agent detection strategy starts with a foundation banks can trust: protected signals, cryptographic device recognition and network intelligence. Personal and malicious agents can look very similar. Our latest advancements in Digital Trust help banks tell them apart, blocking unwanted agents and trusting the ones their customers choose to use.”
Copyright © 2026 RegTech Analyst
Copyright © 2026 RegTech Analyst
