BioCatch has unveiled BioCatch AiRE (Agentic AI Risk Essentials), a capability built to identify AI agents operating inside banking sessions before any funds are transferred.

The new tool lets financial institutions worldwide determine whether an AI agent is acting for a genuine accountholder or on the instructions of a criminal. Banks using the detection capability will also be alerted to signs that an agent’s activity may carry risk.

AiRE draws on a range of signals to flag agentic behaviour, including behavioural irregularities, atypical interaction patterns such as synthetic inputs, the permission profile of the browser, the infrastructure an agent runs on, and whether the agent declares itself. Its coverage extends to cloud-hosted agents such as Muse.

Rather than requiring a fresh deployment, the capability is delivered through integrations that BioCatch customers already have in place. By monitoring sessions on a continuous basis, the company can both spot agents and evaluate how they behave, giving banks the information they need to set policies on which agents deserve trust.

Data from BioCatch’s own customer sessions shows that almost 50% of identified agentic interactions occurred during account opening. The remainder were distributed across logins, account management and transactions, among them electronic fund transfers, wires and faster payments.

BioCatch specialises in preventing fraud, scams and financial crime, using the recognition of human behaviour patterns as the foundation of its approach. Note: the press release provides no further detail on the company, such as its headquarters, size or client base.

The launch comes as concern about AI agents grows on both sides of the banking relationship.

Source link