Banks Warn AI Shopping Agents Outpace Consumer Protections, Pose Fraud and Privacy Risks

Major financial institutions including Bank of America, NatWest, ING and Capital One warned on September 22, 2026, that AI shopping technology is advancing faster than industry standards and consumer safeguards, raising risks of fraud, deception and data privacy breaches, according to a joint bank report. The banks issued principles for developing the technology as OpenAI, Anthropic, Google and Meta push AI shopping tools that let consumers select goods and make payments via AI agents. Retailers are meanwhile competing to influence chatbot recommendations. UK retailer John Lewis said AI-agent-driven searches grew from 0.3% to 2.5% of queries within a year and continue to accelerate. The report said consumers face risks of buying wrong items, overspending and fraud, noting "consumers do not know whether AI will protect their interests." Additional security concerns include AI agents requesting credit card numbers and auto-filling them into websites, or steering users toward weaker payment methods. The banks urged policymakers to mandate clear disclosure when AI is involved in transactions, greater transparency in AI decision-making, stronger data privacy protections, and interoperability ensuring free choice for consumers and merchants.