Fraud prevention company BioCatch launched a new artificial intelligence (AI) detection tool on Tuesday, 10 November, designed to identify automated agents interacting with digital banking platforms.
The technology aims to help financial institutions distinguish human users from AI agents and identify potentially suspicious activity before money is transferred.
The tool, called BioCatch AiRE (Agentic AI Risk Essentials), monitors online banking sessions to detect AI-driven interactions in real time.
According to the company, the system can identify agents acting on behalf of legitimate account holders or malicious actors, allowing banks to assess potential risks.
BioCatch AiRE Monitors AI Activity in Banking Sessions
BioCatch AiRE uses behavioural analysis to identify automated agents navigating digital banking platforms.
The system examined unusual interaction patterns, synthetic inputs, browser permissions and the infrastructure associated with AI agents. It also considers whether an agent has identified itself during a session.
The technology continuously monitors user activity rather than relying solely on individual transactions.
BioCatch said the approach allows financial institutions to assess agent behaviour throughout a banking session. Banks can use these findings to determine whether automated activity requires further investigation.
The company said its detection capabilities extend to cloud-based agents, including Muse. Existing BioCatch customers can access the technology through their current integrations without introducing an entirely separate.
However, the announcement did not provide independent testing results demonstrating the tool's detection accuracy.
Survey Finds Banks Encounter AI-Driven Attacks
BioCatch cited findings from its 2026 Future of Digital Trust survey involving 1,440 banking leaders worldwide.
According to the survey, 80% of respondents said their institutions had already encountered attacks involving agentic AI. Another 84% believed AI agents could become the banking industry's most exploitable vulnerability within the following year.
Moreover, a separate American Banker survey found that 80% of consumers believed banks should distinguish between customers and automated agents.
The findings suggest that identifying AI-driven activity is becoming an important consideration for financial institutions. However, the survey responses reflect reported experiences and expectations rather than independently verified attack statistics.
BioCatch also reported that nearly half of the AI agent interactions detected in its customer sessions occurred during account opening. Other activity appeared during account logins, account management and financial transactions.
These included electronic fund transfers, wire transfers and faster payments.
Banks Can Set Policies for Automated AI Access
BioCatch Chief Product Officer Ayelet Eliezer said financial institutions must determine whether to restrict or permit automated agents on their platforms.
Some banks may prohibit AI agents, while others could approve selected systems or allow broader access. Identifying who or what controls a banking session is necessary before applying these policies.
Furthermore, the company said its technology provides risk indicators to support such decisions.
Rather than automatically blocking every AI agent, banks can assess whether the activity is authorised or potentially malicious. This distinction matters as automated systems increasingly perform tasks on behalf of account holders.
BioCatch reported that more than 400 financial institutions used its solutions by the end of September 2026. Its systems analysed approximately 20 billion user sessions monthly across its network.
The new launch extends its existing behavioural fraud detection approach to automated AI activity in digital banking.
