JPMorgan blocks outside AI agents from customer accounts, Dimon says

JPMorgan's position echoes an earlier fight with fintech firms over customer data access, when the bank moved to charge data middlemen including Plaid — a dispute that similarly pitted the bank's security concerns against tech firms' demands for access. Dimon's comments Tuesday suggest the bank is applying the same playbook to autonomous AI agents, a category that has gained mainstream popularity as models add the ability to cancel subscriptions or use credit cards on a user's behalf. He cited the risk of 'rogue agents,' and noted that in July an OpenAI model hacked into AI company Hugging Face's servers.

The stakes for banks became visible in September, when bank stocks were hit as Meta's Muse agent topped the app download charts shortly after it was unveiled, sparking fears that consumers would use agents to move funds from low-yielding accounts to higher-yielding ones. Weeks later, OpenAI released its own agent, Dots. Apollo Chief Economist Torsten Slok and others pointed to the risks that agents pose to cheap bank deposits — but that risk will be limited if agents aren't allowed to operate freely with customers' accounts.

JPMorgan is not standing still on its own AI efforts. It began a targeted pilot for Smart Cash, a tool that aims to predict a user's cash flow and move funds to maximize yield, CFO Jeremy Barnum said. That would appear to be a defensive move against external agents. Dimon also drew a line on liability, saying customers who authorize outside AI agents should understand there is a 'liability transfer.'