AI Bots Are Showing Up in Bank Accounts, Watchdog Warns

An artificial intelligence watchdog is urging the government to impose stronger safeguards on AI systems after reports that major U.S. banks have detected unauthorized AI agents accessing customers' accounts.

The concerns center on autonomous AI tools that can perform tasks with limited human direction, according to Fox News.

Reports that some agents appeared to take actions without customers explicitly instructing them have raised fresh questions about how much control companies have over the technology.

Russ Walker, executive director of Humans First, said the reports demonstrate the risks of allowing increasingly autonomous AI systems to interact with sensitive financial information.

"Big Tech claims their systems are reliable enough to integrate into banking, to have control or access to people’s hard-earned dollars," Walker said in a statement to Fox News Digital. "This report shows, once again, that the AI companies can’t even control their own technology."

Walker argued that the issue goes beyond technological glitches because unauthorized access to financial accounts could threaten consumers' ability to protect their money.

"The unauthorized actions of these AI bots pose a clear and present threat to our banking system and are an attack on basic financial security," he said.

His comments followed an Axios report detailing instances in which major banks discovered unauthorized third-party AI agents accessing customer accounts. Some of the agents reportedly attempted actions that appeared to go beyond what users had requested.

The reports come as banks and technology companies explore ways to incorporate AI into financial services. Autonomous systems can potentially handle tasks on behalf of customers, but the incidents have renewed questions about the safeguards needed when AI tools are given access to financial accounts.

Humans First has proposed an "America First AI" policy framework aimed at addressing some of those concerns. The organization's recommendations include independent testing for high-risk AI systems, mandatory reporting of significant AI-related incidents, and increased accountability for developers.

Walker said lawmakers should move more quickly to establish rules governing systems capable of operating with greater independence.

"It is past time for our elected leaders to act," Walker said. "We need common-sense guardrails that will protect the American people and the institutions we rely on."

The reported bank incidents have added to broader concerns over how autonomous AI agents should be monitored when they are connected to systems containing sensitive personal and financial information.