BioCatch Launches AiRE to Detect AI Agents in Banking Sessions, Finds Half Occur at Account Opening
The behavioral biometrics firm's new capability distinguishes human-driven from agent-driven sessions before money moves — with early network data pointing to onboarding, not transactions, as the most exposed moment.
Behavioral biometrics provider BioCatch launched AiRE (Agentic AI Risk Essentials) on 10 November, a capability that detects whether a digital banking session is being run by an AI agent — authorized or malicious — before funds move, delivered through the vendor's existing integrations at more than 400 financial institutions.
Detection signals include behavioral anomalies, synthetic input patterns, browser permission profiles, agent infrastructure characteristics and agent self-declaration, with coverage extending to cloud agents. The company's early network data is the notable finding: nearly 50 percent of detected agentic interactions occur during account opening, with the rest spread across login, account management and payment flows. Supporting survey data from 1,440 bank executives found 80 percent saying their institutions had faced agentic AI-based attacks and 84 percent rating agents the industry's biggest vulnerability for the year ahead — all vendor-reported figures.
The launch completes a two-week wave of commercial agent-governance products spanning outright blocking, agent identification, consent enforcement and now continuous session detection — and places the exposed point squarely at identity documents and e-KYC workflows.