AI agents can buy for us, but payments still need to prove who is really in control
Following new Mastercard research predicting that more than one in ten online shoppers will routinely use AI agents to make purchases on their behalf by 2030, Laurence Booth, Group CEO at Trust Payments, comments: “Agentic commerce is rapidly moving to the next frontier of e-commerce. We are already seeing consumers use AI to discover and compare products, however, the much bigger shift comes when they are prepared to give an agent the authority to actually make the purchase.”
“As a principal member of VISA and Mastercard, Trust Payments has a close view of how the payments ecosystem is evolving to support new ways for consumers and merchants to transact. The technology to enable agentic payments is developing quickly, but the harder question is how we prove that a transaction genuinely reflects the consumer’s intent.”
“An AI agent may be authorised to spend on someone’s behalf, but what happens if it buys the wrong product, exceeds the parameters it was given or makes a decision the consumer subsequently disputes? Authentication can tell us that an approved agent initiated a transaction, but that does not necessarily settle questions of accountability or consumer redress.”
“That is particularly important because there is still a distinction between consumers using AI to shop and trusting it to spend their money autonomously. Agentic commerce therefore cannot simply be treated as another checkout channel. Payments providers, card schemes, merchants and regulators will need clearer ways of recording consent, delegated authority and the parameters under which an agent was instructed to act.”
“The real test will be whether the payments ecosystem can demonstrate why that payment was made, what the consumer authorised and where accountability sits when the outcome does not match their intent. Getting that framework right will be fundamental to building the trust agentic commerce needs to scale.”