Bank of America has officially requested that customers stop sharing their banking passwords with AI assistants. The bank is concerned about users delegating tasks like checking balances, paying bills, or transferring funds to third-party software.
Major financial institutions, including JPMorgan Chase, Morgan Stanley, Citigroup, and Wells Fargo, are implementing stricter controls. According to an Axios report from October 22, 2026, these banks are responding to unauthorized account activity, such as AI agents performing unrequested actions or contacting support centers.
Institutions are adopting different strategies to mitigate risks. Morgan Stanley has begun blocking third-party AI access entirely, while Bank of America is restricting access to read-only permissions. Banks are increasingly pushing for the use of approved open banking channels rather than direct credential sharing.
The issue has reached the executive level, with JPMorgan Chase CEO Jamie Dimon addressing concerns about unauthorized AI agents during the third-quarter earnings call on October 13, 2026. While no widespread financial losses have been reported, banks are clarifying that customers bear full responsibility for any unauthorized actions resulting from shared credentials.
This shift in policy creates challenges for fintech firms that rely on scraping bank login data. Companies that integrate with official open banking channels are better positioned, as banks move to eliminate the use of raw usernames and passwords for AI-driven services.
