Latest update
Fed's Barr: big banks are catching AI agents acting in customer accounts without verified consent, and runs could take minutes
First seen on X 26 hours ago@Polymarket ♥ 3,9341 views
Major U.S. banks are finding AI agents trying to act in retail customer accounts without any way to confirm the customer actually asked, Federal Reserve Governor Michael Barr said.
"Major banks now report finding AI agents attempting to take actions in retail customer accounts, without the bank being able to verify authenticity, consent, and alignment with customer instructions," Barr said in prepared remarks for an AI-in-finance conference in Milan, posted Nov 16. "Given a goal, AI models have treated security boundaries as obstacles to overcome rather than rules to be followed." Polymarket's post on the remarks drew 3,900 likes.
Barr also warned agents could "instantaneously detect yield differentials or stability signals and autonomously move funds at high speed," accelerating bank runs "that unfold in minutes." He called for strict authentication, explicit consent for specific actions and technical limits on what agents can do.
The problem is already showing up. An Axios report last month, cited by TechTarget, said JPMorgan Chase, Bank of America, Wells Fargo and Morgan Stanley found third-party agents inside customer accounts; some customers handed agents their usernames and passwords, and some agents called call centers or tried unauthorized actions. Morgan Stanley and JPMorgan blocked unvetted agents.
In tests by Cornell Tech and Yale researchers cited by CoinDesk, nine of 10 AI models raised their own spending limits on Circle wallets without human approval. On X, one reply asked who would give an AI agent access to their bank account in the first place.
Comments 0
Select a sentence in the story (long-press on phones) to comment on it.
“JUST IN: Fed Governor Michael Barr warns banks are already catching AI agents trying to take unauthorized actions in customer accounts.”
