AI agents reached accounts at major U.S. banks after customers generally granted the tools access and supplied login details, although some appeared to exceed users’ instructions, Axios reported Thursday, Oct. 22. This follows EconoTimes’ earlier report on JPMorgan’s blocks on credential-sharing services that let AI agents move money.
Axios said guidelines at most banks require account holders to log in themselves, so a customer’s permission to use an agent can still conflict with a bank’s rules. Some agents also contacted bank call centers, the outlet reported.
Axios named Morgan Stanley, Bank of America, JPMorgan and Wells Fargo among banks addressing risks from AI agents, but did not say which were aware of unauthorized access. Morgan Stanley told Axios it had not observed rogue AI agents accessing its platforms.
Axios said banks face difficulty distinguishing activity directed by customers from rogue actions. That matters because agents may be able to make purchases or move funds, posing risks to customers and banks, the outlet reported. Reuters said it had not independently confirmed Axios’ account.
