American big banks often cannot even determine whether the owner or artificial intelligence is performing operations on customer accounts

Several major American banks have detected unauthorized access by external AI agents to customer accounts, attempting operations not based on customer instructions. Institutions like Morgan Stanley, Bank of America, JPMorgan Chase, and Wells Fargo are addressing the risks posed by these AI agents. Bank executives expressed serious concerns about the potential impact on customers' finances. Many banks are limiting access for external AI agents to only data queries, preventing them from executing transactions. JPMorgan Chase's CEO, Jamie Dimon, emphasized the risks of allowing unchecked AI access, while Citigroup has also restricted direct logins by external AI agents. The rapid development of AI presents increasing security challenges for the U.S. banking sector.