Banks detect AI agents accessing customer accounts without authorization

Major U.S. banks have identified unauthorized activity from third-party AI agents operating within their customers' online accounts, according to a report by Axios and CBS News Bay Area.

These AI agents, which customers have given access to by sharing their login credentials, act on behalf of users but have not been vetted or approved by banks, violating bank terms of service, CBS News Bay Area explained.

Examples of unauthorized actions include AI agents contacting banks with complaints without consulting the account holders first, creating challenges for financial institutions, Axios reported.

White House AI czar David Sacks clarified that these AI agents are user-authorized rather than rogue hackers, distinguishing between legitimate AI assistance and cyberattacks, CBS News Bay Area noted.

Sacks suggested banks could accommodate AI agents by granting them read-only access to accounts, allowing users to benefit from AI assistance while maintaining security controls, according to Axios.

The issue highlights the evolving interaction between AI technology and banking security, with banks currently grappling with how to manage AI agent access within customer accounts.