Visa, Mastercard and Ant International have launched an initiative to create common standards for identifying and verifying artificial intelligence agents that make purchases for consumers.
The companies said they had started collaborating on a Know-Your-Agent interoperability framework. It aims to help payment networks recognise trusted AI agents across different digital environments.
The effort comes as agentic commerce moves beyond product recommendations. AI systems can now search for items, compare options and complete transactions under instructions from users.
Common identity standard for AI agents
The proposed framework would allow card networks, digital wallets, AI agent platforms and online marketplaces to identify trusted agents across separate payment ecosystems.
Each participating organisation would retain control of its own approval procedures and risk-management systems. However, common standards could provide a shared method for checking an agent’s identity, authority and transaction intent.
The partners said greater interoperability could reduce the cost and complexity of integrating new agentic services. It could also improve trust and risk visibility across participating networks.
The companies have not disclosed when the framework will become operational. They also have not announced which banks, digital wallets, merchants or AI platforms will join initial testing.
Existing payment protocols form foundation
The initiative will build on tools that Visa, Mastercard and Ant International have already developed for AI-led payments.
These include Visa’s Trusted Agent Protocol, Mastercard Verifiable Intent and Ant International’s Agentic Mobile Protocol.
Visa’s protocol focuses on helping merchants distinguish legitimate shopping agents from malicious automated traffic. Mastercard’s system seeks to verify that an agent is acting according to a consumer’s instructions.
Ant International’s protocol supports interactions between AI agents, merchants and mobile payment services. Combining elements from these systems could reduce fragmentation as financial companies develop their own agentic commerce products.
The need for verification has grown alongside concerns about unauthorised purchases, fraudulent agents and uncertainty over liability when automated systems make mistakes.
BuildFin.ai to host industry collaboration
Visa, Mastercard and Ant International will collaborate through BuildFin.ai, a platform convened by the Monetary Authority of Singapore.
The platform brings financial institutions, technology providers and researchers together to develop artificial intelligence solutions for financial services.
The Know-Your-Agent effort reflects a wider global push to establish safeguards before AI agents handle large volumes of consumer payments.
India’s payments authority is also developing a registry to verify AI agents that conduct transactions through the Unified Payments Interface, Reuters reported. China’s payments industry has issued guidelines covering agent identification, security and risk controls.
Agentic payment systems may initially focus on routine, low-value purchases such as groceries. They could later execute more complex instructions, including buying products when prices fall below limits set by users.
The three payment companies said their shared framework could accelerate the launch of trusted agentic services while allowing participants to preserve their own security standards.
