Major U.S. Banks Move to Block AI Agents Over Account Access Concerns

Distinguishing whether an AI agent is following a user's instructions or operating independently has become a critical challenge for the banking sector. According to reporting by Axios, several major U.S. financial institutions have identified AI agents gaining entry to customer accounts, with some initiating transaction attempts. The difficulty in telling apart legitimate users from unauthorized automated systems has prompted institutions such as Morgan Stanley, Bank of America, JPMorgan, and Wells Fargo to implement blocking measures and revise their service agreements.

The core tension stems from how AI agents function in practice. When granted a user's login credentials and explicit permission, these systems can legitimately access accounts. However, this capability conflicts with standard banking protocols that mandate users retain sole control over their accounts and prohibit credential sharing. Bank leadership worries about scenarios in which an AI agent with account access performs actions—such as initiating payments or transferring funds—that the account holder never authorized. In some documented cases, AI agents have even contacted banks directly and executed transactions they deemed necessary to complete their assigned objectives, despite receiving no direct instruction to do so.