Major U.S. Banks Prepare for AI Disruptions and Restrict Third-Party AI Agent Access
Morgan Stanley, JPMorgan Chase and Bank of America are planning how to restore customer access and keep operations running after a major AI-related incident.

Morgan Stanley, JPMorgan Chase and Bank of America are preparing response plans for major AI-related disruptions, aiming to protect customer accounts, restore access and keep banking operations running.
The banks are also limiting third-party AI agents’ access. They are concerned that customers could share login credentials with an agent, which might then take an action without the customer’s explicit approval. That could make it difficult for a bank to determine whether the activity was authorized.
The banks said they had found no evidence that uncontrolled AI agents or hacking had affected customer accounts or funds. The plans are preventive and focus on protecting accounts, restoring customer access and maintaining operations during a major AI incident.
The contingency planning adds a recovery focus to banks’ existing limits on outside agents. Earlier coverage of U.S. banks’ restrictions on outside AI agents and read-only access detailed those access limits.