AI Agents Access Customer Accounts, Alarming Major Banks

Jakarta, CNBC Indonesia — Unauthorised third-party AI agents have been detected accessing customer accounts at major banks, including Morgan Stanley, Bank of America, JPMorgan and Wells Fargo.

Sources cited by Axios reported that some agents accessed accounts independently of users’ instructions.

AI agents can access accounts with users’ permission. However, this conflicts with the policies of most banks, which require customers to log in to their accounts themselves.

The issue has raised concerns among bank executives, who worry about agents logging in—particularly if users authorise them to make purchases or transfer funds, Axios reported on Monday (26 October 2026).

The same report said major banks, including Morgan Stanley, JPMorgan, Bank of America and Citi, had moved quickly to restrict agents’ access to their platforms.

A JPMorgan Chase spokesperson said users should not give their usernames and passwords to anyone. A Morgan Stanley spokesperson, meanwhile, said the bank had not found any rogue AI agents accessing its platform.

Some banks have opened access on a read-only basis, allowing AI agents to view financial information but not take any action.

Even so, executives in the banking industry said a balance needs to be found in the use of AI, as customers and banks have markedly different views on the issue.

Customers want to use AI agents in the future. Banks recognise that demand, but consider the risks too high for now.