Mastercard CEO says AI “shopping agents” and machine-to-machine payments are pushing a new layer of commerce infrastructure

Mastercard CEO Michael Miebach said the next wave of commerce will be shaped by artificial intelligence agents that shop and transact on behalf of consumers, and by machine-to-machine payments that occur between connected systems. In Part 2 of an interview published by Yahoo Finance on August 29, Miebach linked those changes to what he described as a new “infrastructure of commerce,” where payments have to work seamlessly with decision-making software rather than only with traditional checkout flows.

Miebach’s comments focused less on a single product launch and more on how AI could alter everyday purchasing. The interview’s framing centered on “AI shopping agents,” meaning software that can research, select, and buy items across merchant sites or channels. The central claim was that as these agents become more capable, they will generate payment activity that looks different from conventional human-driven purchases.

Alongside that, Miebach discussed machine-to-machine payments, referring to payments initiated automatically by systems without direct human interaction at the moment of purchase. While the interview did not lay out specific technical standards or implementation details in the material provided for this review, the underlying point was that payment networks must be built for higher automation, more frequent authorization requests, and smoother handling of transactions that originate from software-to-software workflows.

Mastercard’s role in that future, as Miebach described it, is not only to process transactions but to provide reliability and reach across the commerce stack as new actors enter the buying process. Put differently, when software agents and interconnected devices become purchasing decision-makers, payment rails and authorization systems become part of the operational backbone that those agents rely on to complete transactions.

The interview also emphasized Mastercard’s approach to AI in general terms, describing an outlook tied to utilization rather than experimentation for its own sake. Based on the text available for this review, Miebach did not provide quantified targets, timelines, or specific return metrics, but he suggested that AI capabilities are being brought into the company’s business in ways that support execution across payments.

For Mastercard, the strategic stakes are clear even without new financial disclosures. A shift toward automated buying and ongoing transactions between systems would raise the importance of fraud prevention, risk scoring, and payment authorization efficiency, all areas where large networks typically compete on scale and data. In that context, AI is not just a customer-facing feature, but a lever that can improve how payments are validated and routed under changing commerce patterns.

Still, the discussion in the published post did not provide enough detail, in the material available here, to pinpoint which specific agent behaviors are expected to increase transaction volume, which Mastercard capabilities are being adapted first, or how quickly machine-to-machine payments could become mainstream. It also did not include disclosed partnerships, named pilots, or regulatory filings that would allow readers to verify particular deployments.

Looking ahead, investors and payments watchers may focus on whether Mastercard follows this kind of CEO narrative with concrete product announcements, measurable performance indicators, or partner announcements tied to AI-driven checkout flows and automated payment initiation. The next test of the thesis will be evidence that software agents and machine-to-machine transaction models translate into expanded usage and stronger economics for the payments network, not only into improved internal efficiency.