A coalition of major financial institutions has issued a warning regarding the deployment of artificial intelligence agents for online shopping, highlighting significant vulnerabilities related to scams, fraud, and data privacy for consumers.
This alert, detailed in a recent report authored by a consortium of banks—including Bank of America, Capital One, the UK’s NatWest, New Zealand’s ASB Bank, and Commonwealth Bank of Australia—cautions that while the potential for AI shopping agents is promising, the technology is currently advancing faster than consumer protections and industry standards.
“Consumers remain uncertain whether AI will act in their best interests,” the report stated. “They worry that AI agents might purchase incorrect items, overspend, or—more critically—fall victim to scams and fraud, losing their money. Furthermore, consumers are unclear about whether they will be protected or whom to contact if something goes wrong.”
Technology firms are rapidly deploying increasingly sophisticated AI agents to consumers, utilizing them as shopping tools capable of tracking products and executing purchases on behalf of the user.
A man looks at bills next to a computer in his home. (Getty Images)
British retailer John Lewis reported in September that searches originating from AI agents surged from 0.3% last year to 2.5% this year, reflecting a growing share of traffic driven by these automated systems.
The banks’ report flagged several risks associated with AI agents, such as requesting customers’ card details and inputting them directly into websites, or redirecting consumers toward payment methods that offer weaker consumer protections.
In response, the banks intend to present proposals to policymakers focused on consumer protection. These proposals include mandating disclosure whenever an AI agent participates in a transaction, enhancing transparency regarding how AI agents make decisions, and implementing safeguards to secure customer data.
The report emphasized that consumers and merchants should retain the freedom to choose their preferred AI-powered e-commerce services, while ensuring different systems remain interoperable.
Agentic AI systems can take actions and interact with other systems with a degree of user-granted autonomy. Unlike conversational AI chatbots that merely respond to prompts or conduct research, agentic AI is capable of actively performing tasks.
Banks warned that agentic AI could create risks for consumers if not held in check. (iStock)
Agentic AI apps can make purchases and fill out forms at the direction of their user. (Getty )