Even if broad consumer use of agentic shopping is not yet prevalent, much of the infrastructure needed to support it is already being developed. But as the technology advances, a group of leading banks has raised concerns that industry standards and consumer protections have not kept pace.
According to Reuters, the consortium—which includes Bank of America, Capital One, NatWest, and ING—has begun establishing principles for safeguarding agentic commerce.
In a report, the group highlighted the potential of agentic commerce while noting that many consumers may be willing to embrace it. However, widespread use of AI agents within today’s e-commerce environment could also increase the risk of scams, fraud, and privacy breaches, and potentially create a significant barrier to adoption.
The report noted that consumers are uncertain about whether AI agents will act in their best interests. Concerns include agents making unintended purchases, spending more than authorized, or exposing consumers to scams and fraud.
Echoed and Amplified
These concerns have become increasingly prominent as agentic commerce move closer to broader adoption. Industry efforts are underway to address some of these gaps, including initiatives focused on creating common frameworks for agent-to-agent payments and commerce.
Still, important questions remain around issues such as how a consumer’s intent can be verified and how responsibility for an AI agent is determined.
The banks’ report identified additional risks. Privacy and security concerns could arise, for example, if AI agents are able to request customers’ card details and enter them directly into websites. There are also questions about how to prevent agents from steering consumers toward payment methods that don’t have that same protections as more established options.
The Breakneck Pace of Evolution
Despite these concerns, the march toward agentic commerce continues. Major AI organizations, including OpenAI, Anthropic, and Google, are building new platforms and exploring ways to incorporate AI agents into the retail experience.
Agentic payments may not yet be a mainstream retail experience, but the rapid evolution of generative AI suggests that adoption could develop quickly. The pace of change in AI over the past few years illustrates how quickly expectations and technology can shift.
To get ahead of these challenges, the group of banks plans to discuss a series of proposals with regulators to iron out the safeguards and privacy mechanisms before agentic commerce gains broader traction.
