Major U.S. banks restrict AI agents using shared customer logins

Major U.S. banks are blocking or limiting third-party AI agents that reach accounts after customers share credentials, with no confirmed bank-system breach.

An Axios report, which Reuters said it could not immediately verify, says major U.S. banks are finding third-party AI agents inside customer accounts after customers share usernames and passwords. That practice conflicts with rules requiring customers to log in themselves. GBHackers, citing Axios, frames the issue as customer authorization rather than a breach of bank systems or stolen credentials; Axios described no confirmed widespread theft. Morgan Stanley is blocking such agents, while JPMorgan, Bank of America, Citi and others are restricting them, generally toward read-only access. A Morgan Stanley spokesperson said it had not observed rogue agents on its platforms, and restrictions do not confirm every named bank saw unauthorized access.