Unified AI Agent Identification System in Development

Ant International, Mastercard, and Visa began working together on a system designed to standardize the identification and onboarding of AI agents across payment networks and digital wallets.
Ant International, a global provider of digital payment and financial technology solutions, joined Mastercard and Visa to develop the Know-Your-Agent (KYA) framework. The initiative targets card networks, digital wallets, AI agent platforms, and marketplaces and aims to simplify the identification of autonomous agents operating across different payment ecosystems.
The project aims to establish common trust signals that can be recognized across different segments of the payment infrastructure. For platforms and marketplaces, this approach could eliminate the need to undergo similar onboarding and verification procedures separately for each participating network. Payment networks will retain their own verification, risk management, and decision-making mechanisms. The approach is also intended to reduce integration complexity and the cost of launching new agentic commerce services.
For more on agentic commerce and how AI agents are changing the customer journey, merchant operations, and competition in e-commerce, read this special report from CP Media.
The project participants already have their own technology protocols for these operations. Visa developed the Trusted Agent Protocol, Mastercard created Verifiable Intent, and Ant International developed the Agentic Mobile Protocol. The companies will now explore ways to make these protocols interoperable based on a common set of KYA principles.
The KYA architecture is expected to incorporate 3 core mechanisms:
- End-to-end operator traceability: Each AI agent will be linked to a verified operator, cardholder, company, or other organization, making it possible to clearly establish accountability for its actions.
- Common certification requirements: AI agents will be assessed for compliance with security requirements and established behavioral criteria.
- Continuous transaction monitoring: AI agent status will be continuously assessed using identity and transaction signals to manage risk and ensure certifications remain current.
The companies said the need for this infrastructure stems from AI agents moving beyond recommending products and services to making purchases independently on users’ behalf. According to an estimate cited by the companies, AI agents could facilitate between $3 trillion and $5 trillion in global consumer purchases by 2030.
A separate area of collaboration among Ant International, Mastercard, and Visa will be developed in Singapore through BuildFin.ai, an industry platform created with the involvement of the Monetary Authority of Singapore (MAS). The companies plan to develop common approaches to AI agent verification, allocation of responsibility, and risk management across payment ecosystems.
Previously, Ant International participated in testing Agent Pay for Machines (AP4M), an automated settlement platform developed by Mastercard.

