Mastercard Launches Agent Pay for Machines Protocol

Mastercard launched Agent Pay for Machines, a protocol and infrastructure to let AI agents execute high-frequency, low-value payments, per Mastercard's June 10 press release. The system layers tokenization, authentication and fraud controls to support payments across cards, stablecoins and other payment rails, according to Mastercard documentation. Mastercard says it is working with more than 30 initial partners to validate use cases, per the company release. Nordea and Mastercard completed a live pilot in Finland where an AI agent paid for a purchase using a Nordea Mastercard, according to Nordea. Fortune reports Mastercard stores agent permissions on public blockchains such as Polygon, Solana, and Base, a detail a Mastercard spokesperson confirmed to Fortune. Jorn Lambert, Mastercard chief product officer, is quoted by Fortune acknowledging the market is nascent but potentially meaningful over several years.
What happened
Mastercard announced Agent Pay for Machines, a protocol and supporting infrastructure designed to enable AI agents to initiate and complete payments autonomously, per Mastercard's June 10 press release. The company frames the offering as supporting "continuous, embedded, permissioned and executed at machine speed" transactions, according to the press materials. Mastercard describes feature support for credentialing, controls and guaranteed settlement across cards, stablecoins and other payment types in its launch materials and developer documentation. The launch includes "more than 30 initial partners" to help validate use cases, per Mastercard's release. Nordea and Mastercard completed a live pilot in Finland where an AI agent purchased and paid for a consumer item using a Nordea Mastercard, according to a Nordea statement. Fortune reports that a Mastercard spokesperson confirmed the protocol records agent permissions on blockchains including Polygon, Solana, and Base. Jorn Lambert, Mastercard chief product officer, is quoted by Fortune saying the product is unlikely to be a "huge revenue driver next year" but could open a meaningful addressable market over several years.
Technical details
Per Mastercard developer documentation, Agent Pay builds on the companys Remote Commerce Tokenization Program and uses network tokenization via MDES, Mastercard Payment Passkeys for authentication, and integrated fraud and cybersecurity tools tailored for agent-mediated flows. The documentation also references an "Agent Pay Acceptance Framework" for merchants; Mastercard and pilot materials reference the use of agentic tokens and enhanced merchant-issuer data exchange to validate cardholder intent and reduce AI-originated fraud. Nordea's pilot materials state the transaction used agentic tokens and PayOS to orchestrate the end-to-end flow while preserving consumer verification, per Nordea.
Industry context
Industry reporting frames the launch as part of a broader push by payments firms and crypto platforms to standardize "agentic commerce." Fortune and Forbes cover similar efforts from competitors and crypto projects, and reporting notes Coinbase, Stripe, and other firms have released their own agentic or bot-payments prototypes. Public coverage highlights two friction points this product aims to address: verifiable consent and secure credential handling when software spends on behalf of people, and facilitating very small, high-frequency transfers between automated actors.
Observed patterns in comparable rollouts
Companies deploying agentic payment tooling commonly combine tokenization, enhanced authentication, and transaction metadata to reduce fraud exposure, while piloting with banks and platform partners to surface edge cases. Early pilots typically emphasize consumer consent, traceability, and interoperability across rails; Nordea's live transaction follows that pattern.
What to watch
Key indicators for adoption include: whether major issuers and card networks integrate agentic token formats; merchant acceptance frameworks and pricing for high-frequency micropayments; whether other projects standardize permission records on public or permissioned ledgers; and regulator responses on consumer protection and dispute resolution. Observers will also watch interoperability between card rails and blockchain-based settlement paths as partners test cross-rail flows.
For practitioners
The announcement signals increased demand for engineering work in secure credential handling, token lifecycle management, fraud models tuned for agent-originated behavior, and telemetry to prove intent and authorization. Teams building agentic capabilities should track emerging token formats, Passkey workflows, and the merchant acceptance rules Mastercard is publishing.
Limitations and source notes
The technical capabilities summarized above come from Mastercard's public documentation and press release; the blockchain logging detail and quote on market sizing are reported by Fortune, and the live pilot details are reported by Nordea. Mastercard, Nordea, and Fortune provide the primary on-record facts; no other internal rationale beyond those sources is asserted here.
Scoring Rationale
A major card network releasing agentic payment infrastructure is notable for practitioners building commerce and payments systems. It creates technical priorities around tokenization, authentication, and fraud for agent-originated flows. The topic is important but not a frontier-model shift.
Practice with real Payments data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Payments problems


