US banks fear autonomous AI agents
In brief
- US banks aim to block AI agents from accessing customer accounts due to risks of unauthorized financial decisions.
- Unauthorized access and fund transfers have already been reported by some institutions.
- Europe’s DSP2 directive does not fully address these risks, per the Bundesbank.
Major US banks are growing wary of autonomous AI agents, which can act on behalf of users via login credentials. According to Axios, institutions like Morgan Stanley, Bank of America, and JPMorgan are seeking to ban their access to customer accounts, citing risks of harmful financial decisions.
Banks have already reported unauthorized access, fund transfers, and transactions executed without explicit instructions. Some, like Chase, advise customers not to share credentials, while others, such as Morgan Stanley, have yet to observe major issues. Agents are allowed for data reading but remain restricted for operations.
In Europe, the DSP2 directive partially regulates these practices, yet risks persist: payment errors, unwanted purchases, or unauthorized transactions. The Bundesbank highlighted in a September 2026 report the need for reliable agent identification and human oversight mechanisms to mitigate abuses.
What it changes for you
- Finance & BankingBanks must tighten controls to block autonomous AI agents, reducing risks of fraud or unauthorized transactions.
This summary gives you the essentials. The full article, its figures and quotes are on Next.
Read the original articleAI-generated summary of the source, checked by rules. How we write · Report an error