Visa, Mastercard and Ant International said on September 10, 2026 that they will build a shared “Know Your Agent” (KYA) framework for identifying AI agents that make payments. The goal, according to TNGlobal, is to link every paying agent to a validated operator, cardholder or business, whichever network the payment runs on.
What changed
Each company already runs its own system for recognizing agents. PYMNTS reports that Visa uses its Trusted Agent Protocol, Mastercard uses Verifiable Intent, and Ant International uses its Agentic Mobile Protocol. The three now plan to agree on common principles so an agent that is trusted on one network sends signals the others can read.
TNGlobal says the framework covers three areas:
- Operator traceability across networks, linking an agent to the person, company or organization that runs it.
- Shared certification requirements for agents.
- Continuous monitoring of agent transactions.
The work was announced in Singapore through BuildFin.ai, a platform convened by the Monetary Authority of Singapore, per TNGlobal. The companies describe it as a framework under development, not a finished technical standard.
What it means for you
Today you can’t easily tell whether an AI agent that buys something for you is recognized by your card network. A shared identity layer is meant to answer that: if an agent’s purchase can be traced to you or to the company that runs it, a disputed charge has a clearer paper trail.
For small merchants, PYMNTS notes the pitch is fewer separate integrations and more consistent trust signals, which could mean fewer legitimate agent orders declined as fraud.
Nothing changes at checkout yet. Until standards ship, the practical protections are the ones you control: spending limits and approvals on the agent and on your card. Our guide to spending limits for AI agents shows where to set them, and the agentic commerce protocol tracker lists which protocols each company supports.
