Major banks express concern over AI agents and increased fraud risks

A consortium of global financial institutions, including Bank of America and ING Group, has published a principles paper titled 'Building Trust in Agentic Commerce.' The banks warn that the rise of autonomous AI agents capable of making purchases on behalf of consumers introduces significant risks, including potential scams, fraud, and unauthorized spending. The institutions argue that current AI development lacks sufficient safeguards, leaving consumers vulnerable to social engineering and merchants exposed to increased chargebacks. The paper outlines five core principles—Transparency, Safety, Privacy, Choice, and Interoperability—that the banks urge AI developers to adopt. This warning coincides with recent security concerns regarding Meta's Muse AI, which Amazon has blocked from its platform due to vulnerability issues. The banks emphasize that without clear liability frameworks and safety standards, the integration of agentic AI into commerce could undermine consumer trust and financial security.