Mastercard and Google Launch Verifiable Intent for AI Payments

Mastercard and Google have introduced Verifiable Intent, an open-source cryptographic framework designed to establish a tamper-resistant record linking a consumer’s verified identity, their specific instructions to an AI agent, and the resulting transaction outcome. The framework targets the trust problem posed by agentic commerce, where autonomous AI systems execute purchases on behalf of consumers without direct human involvement at the point of transaction.

Verifiable Intent is built on standards from the FIDO Alliance, EMVCo, the Internet Engineering Task Force, and the World Wide Web Consortium. A Selective Disclosure mechanism ensures each transaction party receives only the minimum information necessary to verify authorization or resolve a dispute. The framework is protocol-agnostic and is designed to be compatible with Google’s Agent Payments Protocol and Universal Commerce Protocol.

Mastercard is open-sourcing the specification and an initial reference implementation at verifiableintent.dev and on GitHub. Integration into Mastercard Agent Pay’s intent APIs is planned for the coming months. Partners supporting the initiative include Fiserv, IBM, Checkout.com, Basis Theory, and Getnet.

Pablo Fourez, Chief Digital Officer at Mastercard, framed the framework’s purpose directly: “As autonomy increases, trust cannot be implied. It must be proven.” Stavan Parikh, VP and General Manager of Payments at Google, described Verifiable Intent as “a natural accelerator for scaling agentic commerce” when combined with Google’s Agent Payments Protocol.

The framework draws on FIDO Alliance’s dedicated payments authentication work, formalized through a Payments Working Group the alliance launched in 2025. The biometric layer enables a human owner to establish a verifiable intent to pay before an AI agent can finalize a purchase, creating a cryptographic audit trail accessible to merchants, issuers, and dispute systems.

Verifiable Intent extends Mastercard’s wider push to eliminate passwords by 2030, applying biometric-first trust logic to autonomous AI transactions. The framework enters an active market: Wink and Vouched unveiled parallel biometric human-authorization systems for AI agents in late March 2026, while World ID released an Agent Kit linking iris-verified identity to AI agent actions through zero-knowledge proofs.

Sources: Mastercard, FIDO Alliance, PYMNTS

–

By the ID Tech Editorial Team