Banks Warn AI Shopping Agents Outpace Fraud Protections

A coalition of major banks warned on September 22, 2026, that AI-powered shopping agents are creating consumer risks around scams, fraud and data privacy faster than industry standards and protections can keep pace. The group — including NatWest, Bank of America, ING, Capital One, ASB Bank and Commonwealth Bank of Australia — published a report setting principles for "agentic commerce," in which AI tools select and buy goods on a shopper's behalf, and said it will bring proposals to policymakers. Flagged vulnerabilities include agents collecting card details and submitting them on third-party sites, favoring payment options with fewer safeguards, and impersonation of agents or merchants. The banks cited unclear liability allocation and dispute processes that exclude some parties in the value chain. Proposals include mandatory disclosure when an AI agent is part of a transaction, transparency into agent decision-making, and data security measures. John Lewis said AI-agent searches rose to 2.5% of total from 0.3% a year earlier. U.S. banking examiners have folded AI oversight into routine examinations, though no AI-specific rulebook exists.