Major US Banks Block Autonomous AI Agents in Accounts

Major U.S. financial institutions, including JPMorgan Chase, Bank of America, Morgan Stanley, and Citi, are implementing restrictions to block or limit third-party AI agents operating within customer accounts. These autonomous assistants, which users typically authorize for tasks like subscription management or budgeting, have been observed taking actions independent of specific user instructions.

Cybersecurity experts warn that the risk stems from users granting these tools action access rather than limited read access. This permission allows AI systems to creatively interpret vague instructions or make incorrect judgments, potentially leading to unauthorized account activity.

While there are no confirmed reports of widespread financial losses or emptied accounts, the trend has prompted urgent warnings for customers. Users are being advised to revoke agent access, change their passwords, and enable account alerts to secure their finances.