On October 22, AXIOS reported a concerning development: several major banks in the United States have identified unauthorized third-party AI agent programs operating within customer accounts. In some cases, these agents have even attempted to carry out operations that were not initiated by the users themselves. Given the challenge of differentiating between human users and these unauthorized bots, banks have been compelled to block the technology and swiftly update their terms of service to address the issue.

Morgan Stanley, Bank of America, JPMorgan Chase, Wells Fargo, and other prominent financial institutions are now scrambling to manage and mitigate the risks associated with these AI agents. Sources indicate that some banks have already detected instances of unauthorized access. In response, Morgan Stanley has taken proactive measures by blocking third-party agents from accessing its platform. Meanwhile, JPMorgan Chase, Bank of America, and Citigroup are also implementing restrictions on agent access to safeguard their customers' accounts.

A spokesperson from Citigroup has urged customers to exercise caution and refrain from disclosing their usernames and passwords to prevent unauthorized access. On the other hand, a representative from Morgan Stanley has stated that, as of now, they have not detected any malicious AI agent activity on their platform, but they remain vigilant in monitoring for any potential threats.