Banks Warn AI Shopping Agents Pose Fraud and Privacy Risks

Several banks, including NatWest and Bank of America, warned in a new report that letting AI agents shop online for consumers could raise the risks of scams, fraud and data privacy leaks. OpenAI, Anthropic, Google and Meta are pushing AI chatbots into shopping, hoping consumers will let AI agents pick products and place orders directly. Retailers are also competing to influence chatbot product recommendations. UK retailer John Lewis said in September that searches from AI agents rose from 0.3% a year ago to 2.5%, and the growth is accelerating.

The report says consumers are very interested in AI agent shopping, but technology is moving faster than industry standards and consumer protection rules. Consumers are not sure whether AI truly acts in their interest. They worry AI may buy the wrong items, spend too much, or lose money to scams and fraud. They also do not know if they will be protected or who to turn to when problems arise.

Risks listed in the report include AI agents asking consumers for bank card details and entering them directly into shopping sites, or steering users toward payment methods with weaker consumer protection. The banks plan to discuss proposals with policymakers, including requiring clear disclosure of whether AI agents are involved in transactions, improving transparency in decision-making, and setting up customer data protection mechanisms. The report also says consumers and merchants should be free to choose AI e-commerce services, and different systems should be interoperable.