JPMorgan Chase Shifts AI Agent Liability to Customers
JPMorgan Chase CEO Jamie Dimon warns of security risks from AI agents and signals that customers will bear liability for autonomous transaction errors.
Financial institutions are rapidly deploying agentic AI to automate banking transactions, leading to a critical debate over authentication and liability. Jamie Dimon, CEO of JPMorgan Chase & Co., warned that granting autonomous AI agents access to bank accounts introduces significant security risks. He signaled that the bank intends to shift the responsibility for the misuse of these systems onto the customers, a move that could influence terms of service across the U.S. banking sector and impact fintech startups developing autonomous payment software.
JPMorgan Chase has already integrated nearly 1,000 AI use cases across risk, fraud, and marketing, resulting in staff reductions of 30% to 40% in some positions. Other institutions are also adopting the technology: BNY is treating AI agents as digital employees with specific identities, and LHV Bank has deployed agentic AI for customer emails. In Denmark, Danske Bank and Mastercard recently executed the country's first AI-driven payment, where an agent handled a booking and payment on a customer's behalf.
These developments have prompted regulators to question the legal framework of agentic commerce. At the Sibos conference, Federal Reserve Board member Christopher Waller highlighted the trust problem regarding fraud and authentication, specifically questioning who is held accountable when an AI agent makes an incorrect purchase.
