Six major banks published joint principles on Sept. 22 proposing that every party to an agentic commerce purchase (a transaction carried out by an AI shopping agent on a consumer's behalf) know when an AI agent is involved in the transaction and whom it is acting for.

The paper, titled "Building Trust in Agentic Commerce," was co-authored by NatWest Group, Bank of America, Capital One, ING Group, Commonwealth Bank of Australia, and New Zealand's ASB Bank, a lineup spanning banking markets in North America, Europe, and Asia-Pacific. The principles are voluntary rather than binding. The six banks plan to consult policymakers on the proposals and to publish a second paper on how to implement them; the full announcement is available in NatWest Group's press release.

Transparency tops a list of five principles

The disclosure proposal falls under transparency, the first of five principles. The other four are safety, privacy and data, choice, and interoperability.

Transparency also extends to how shopping agents rank options and make decisions, including sponsored options. The paper warns that agents may promote products or payment methods that give their providers a larger cut, such as higher commissions, even if they are not the best value for the consumer.

On safety, the banks want consumers to control the authority they grant to agents, including the ability to view and manage delegations. Agents should enter payment credentials in a secure, auditable manner. Every party should be involved in the dispute process, and liability should sit where the risk or error was introduced.

The privacy and data principle asks providers to keep auditable records of consumer instructions and intent, touch only the data they need, and let consumer and merchant consent determine any further use.

Consumers and merchants, meanwhile, should be free to choose agentic commerce services without unreasonable constraints.

"But for that potential to be realised, customers need to trust that they remain in control of how payments are made and that their money is safe," NatWest Chief Payments Officer Mark Brant said.

Card details keyed straight into websites top the unsafe list

"Consumers are unclear if AI agents will act in their interests," the paper said. Consumers worry that agents may buy the wrong thing, spend too much, or lose money to scams and fraud, and they do not know whom to turn to when things go wrong.

The paper identifies unsafe practices among some providers, such as requesting consumer card details and then keying them into websites, and prioritizing payment methods that offer less protection. It also flags attackers who compromise or impersonate AI agents and merchants. Merchants, for their part, fear that disputes and chargebacks may rise for reasons outside their control.

Agentic commerce has already moved from closed trials to live pilots and early scaling, and those developments are outpacing formal governance, the banks said. As reported by Cryptopolitan, Robinhood rolled out credit card spending for AI agents in May while most companies were still stuck in pilot mode.

The six banks invited other banks and payments firms to apply the principles and help build a blueprint for implementing them. Because the principles carry no binding force, their reach will rest on firms choosing to adopt them; the planned consultations with policymakers and the follow-up paper on implementation are the next checkpoints to watch.