
Bank of America is begging customers not to hand their passwords to AI assistants
AI agents are getting smarter, but banks aren't ready to hand them the vault keys.
via Android AuthorityReported by 2 outlets
The story so far
AI summary of the reporting below · Updated · How we work
Bank of America has sent an email alert titled "Protecting your finances in the Al World" warning customers never to give their banking usernames or passwords to an outside service or person, including any AI-enabled tool or autonomous agent acting on their behalf. A spokesperson for the financial institution confirmed the legitimacy of the message after copies surfaced on social media. [1][2]
The lender cautioned that users who voluntarily provide their credentials to AI assistants remain responsible for any resulting account activity, subject to applicable consumer protections. Beyond unintended financial transactions, the practice risks triggering automated fraud lockouts or exposing sensitive login information to cybercriminals if an external assistant suffers a security breach. [1][2]
The alert comes as financial institutions and payment processors grapple with autonomous tools like Gemini, Claude, and ChatGPT managing personal errands. While Bank of America highlighted its own internal investments in AI services, traditional banks are still evaluating how to safely support emerging features such as agentic payments without exposing customer balances to unauthorized third-party access. [1][2]

AI agents are getting smarter, but banks aren't ready to hand them the vault keys.
via Android Authority
The banking giant reminds customers that "you're responsible for activity" resulting from your AI agents' access.
via TechCrunch