AI Payments: Visa, Mastercard, and Ant Launch Common Trust Framework

The project, named “Know-Your-Agent” (KYA), is designed to enable card networks, digital wallets, agent platforms, and online marketplaces to recognize trusted AI agents across different payment ecosystems while maintaining their own approval and risk management procedures.

The framework builds on separate initiatives already developed by the three companies: Visa’s Trusted Agent Protocol, Mastercard’s Verifiable Intent, and Ant International’s Agentic Mobile Protocol. The stated goal is to align these approaches around common principles rather than impose a single system. Mastercard is also developing other digital payment applications, particularly in conversational services.

The companies outline three key areas of focus: linking each agent to an operator, cardholder, or validated organization; defining common certification requirements regarding security and behavior; and ensuring ongoing transaction monitoring based on identity and risk signals.

The collaboration will be conducted under BuildFin.ai, a platform that brings together financial institutions, technology providers, and researchers, driven by the Monetary Authority of Singapore. It will specifically address agent verification, accountability, and risk management in financial services.

AI agents are expected to move directly to making purchases

The payments sector is preparing for a shift where AI assistants will not only recommend products or compare offers but can also initiate transactions on behalf of users, within defined permissions. This trend aligns with the emergence of agents capable of executing end-to-end tasks, including payments.

In their joint statement, Ant International, Mastercard, and Visa reference a McKinsey projection that AI agents could facilitate between $3,000 billion and $5,000 billion in global consumer commerce by 2030. The three companies present the identification of agents and the traceability of their actions as necessary conditions for this growth.

The partners believe that an interoperable framework could reduce repeated identity verifications for agents and integration costs for new services while maintaining trust and risk controls specific to each network.

A study published by Visa on September 9 highlights the trust issue: 23% of surveyed American consumers say they trust generative AI to manage payments on their behalf. Visa claims this percentage increases when traditional payment players are involved in the transaction.

An initiative led by Singapore

The announced work focuses on the interoperability of trust signals between cards and digital wallets. It will not replace the verification and risk management procedures specific to each network.