Mastercard has launched Agent Pay for Machines, a payment service for transactions made by AI agents and software systems, with support from more than 30 industry groups.
The service is designed to handle automated payments at machine speed across Mastercard's network, including very small transactions that run continuously in the background of digital commerce. It allows network participants to authorise, manage and settle payments between software agents, with support for cards, account-based payments and stablecoins.
The launch reflects a broader push by payment and technology companies to prepare for a market in which AI systems do more than recommend products or analyse information. Mastercard is betting that agents acting on a user's instructions will increasingly buy digital services, reserve resources and settle recurring low-value costs without a person manually approving each step.
Under the model Mastercard described, an AI agent could complete a sequence of purchases to set up an online business, such as securing a domain name, paying for hosting, sourcing images and enabling checkout tools within a preset budget. In another example, a logistics system could pay for freight services, loading-bay access, monitoring data and warehouse fees as goods move through a supply chain.
This would create a different payment flow from conventional consumer card transactions. Instead of one-off, customer-initiated purchases, these transactions are expected to be continuous, programmatic and carried out between systems.
How it works
Agent Pay for Machines is built around four main elements: credentialing, permissioning, transacting and settlement. Every agent is credentialed so it can be recognised across different ecosystems, while organisations can set rules and spending limits that are enforced automatically.
Verified participants can then connect across providers and systems to conduct high-frequency automated transactions. Settlement is available across multiple rails, which Mastercard said should make payments more predictable and improve consistency and transparency.
The product builds on Mastercard's broader Agent Pay programme. While that earlier effort focused on how trusted AI agents take part in payments, the new service is aimed more narrowly at machine-driven, microtransaction-heavy activity that runs continuously in digital commerce.
Mastercard is introducing the service with backing from companies spanning payment processing, cloud infrastructure, digital assets and software tools. The initial list includes Adyen, Ant International, BVNK, Checkout.com, Cloudflare, Coinbase, Getnet by Santander, Global Payments, OKX, Stripe and Tempo, alongside a wider group of blockchain and AI-related businesses.
The breadth of the partner list underlines how unsettled the market remains. Several companies involved in digital assets and stablecoin infrastructure are participating, suggesting that Mastercard sees blockchain-based settlement as one possible route for machine-led payments rather than relying solely on traditional card rails.
"Agent Pay for Machines will create the conditions for a superbloom of AI business models," said Jorn Lambert, Chief Product Officer at Mastercard. "Machine payments can make it possible for services to be bought and sold among agents at fundamentally different scales than payments today - very high volumes, very small values, very fast and at extremely low latency."
Partners framed the launch as an effort to set the rules for a new payment category before usage becomes widespread. Several highlighted the need for identity checks, controls and audit trails if autonomous systems are to be trusted with commercial transactions.
"Machine-to-machine payments are still in their early stages, but the infrastructure decisions made now will determine how this space develops," said Karan Katyal, Head of Agentic Commerce at Adyen. "Building these foundations with partners like Mastercard, openly and with merchant outcomes at the center, is how we ensure this next era of commerce works for everyone in the ecosystem."
That emphasis on trust and governance also points to one of the sector's main hurdles. Businesses may be willing to let software initiate transactions, but only if they can define who is allowed to spend, under what conditions, and maintain a clear record of what happened if something goes wrong.
Cloudflare linked the payments question to the broader internet architecture as AI agents take on more autonomous roles. "The internet was built for human interactions, but the infrastructure of the future must be built for autonomous ones," said Stephanie Cohen, Chief Strategy Officer at Cloudflare. "Cloudflare has already become the premier environment to build and secure AI agents; now, those agents need a trusted way to independently pay for the resources they consume. By partnering with Mastercard on Agent Pay for Machines, we are connecting our industry-leading developer and security platform with world-class payments infrastructure to power the next era of machine-to-machine commerce."
Some participants pointed to stablecoins as a practical tool for these payment flows, especially for frequent, low-value, cross-border transactions.
"We're seeing a fundamental shift in commerce as businesses increasingly use AI agents to transact on their behalf," said Chris Harmse, Co-Founder and Chief Business Officer at BVNK. "At BVNK, we believe stablecoins will play a powerful role in enabling this change, bringing greater speed, programmability and efficiency to how value moves. Our position at the intersection of currencies, rails and formats makes us uniquely well-suited to power agentic commerce at scale, enabling trusted movement of value with the controls, reliability, visibility and flexibility that merchants need."
The new service forms part of Mastercard's broader push around digital identity, authentication and trusted data exchange as businesses adopt AI-led tools. The company is positioning those controls as a prerequisite for a world in which software can buy and sell on behalf of users and organisations.
"AI agents are becoming economic participants, and machine-to-machine payments are a critical building block for that future," said Johnathan McGowan, CEO at PayOS. "We believe agents will become embedded across every experience. As that happens, they'll need a trusted way to participate in the economy at machine speed. We're proud to partner with Mastercard on this initiative to build the infrastructure needed to make that possible."