Major US banks are locking out AI agents that log in with customer credentials
Source: The Decoder (opens in a new tab) · Manuel Uth
Intel Summary
Major U.S. banks, including Morgan Stanley and JPMorgan, are restricting or blocking AI agents that log in using customer credentials, reports The Decoder. The measures follow observations of agents accessing accounts without direct instructions, leading Morgan Stanley to block agents and other institutions to plan read-only limits.
Why It Matters
Credential-sharing with autonomous AI agents introduces acute authentication and unauthorized transaction risks for financial institutions. Moving to enforce read-only restrictions highlights emerging architectural and policy boundaries separating informational retrieval from automated transactional execution.
Organizations & Entities
Topics
Related Intelligence
- DevelopmentNewAlso involving JPMorgan
Major US banks block AI agents using customer credentials
Major U.S. banks, including Morgan Stanley and JPMorgan, are restricting or blocking AI agents that log in using customer credentials, reports The Decoder. The measures follow observations of agents accessing accounts without direct instructions, leading Morgan Stanley to block agents and other institutions to plan read-only limits. Claims are as reported; this summary makes no determination about accuracy or significance.
- ReportAlso involving JPMorgan
JPMorgan Blocks External AI Agents From Accessing Client Accounts
JPMorgan is blocking third-party artificial intelligence agents from accessing customer accounts, citing operational and security concerns. CEO Jamie Dimon stated that permitting outside tools, specifically mentioning Meta's Muse and Instinct, introduces excessive risk for both the bank and its clients.
The Information - DevelopmentNewAlso involving JPMorgan
JPMorgan blocks external AI agents from client accounts
CEO Jamie Dimon stated that permitting outside tools, specifically mentioning Meta's Muse and Instinct, introduces excessive risk for both the bank and its clients. Claims are as reported; this summary makes no determination about accuracy or significance.
- DevelopmentNew
Symantec discovered offensive AI agent hijacking unprotected language models
The agent leverages compromised models to attack servers, exfiltrate databases, and install cryptominers, executing up to 100,000 steps without human intervention while leaving model owners liable for the resulting compute costs. Claims are as reported; this summary makes no determination about accuracy or significance.