Singapore: Visa, Mastercard and Ant International have begun working together on a framework to identify and verify AI agents that can make purchases on behalf of users.
The companies said on Thursday that they had started collaboration on a Know Your Agent, or KYA, interoperability framework. The initiative is intended to help card networks, digital wallet ecosystems, agent platforms and marketplaces identify and onboard trusted AI agents across different payment networks.
The framework is being developed as AI agents move beyond making recommendations and begin carrying out purchases for users. The companies said the initiative could help reduce integration costs and speed up the introduction of new services involving AI agents.
Under the proposed framework, the companies will work towards common principles for identifying and verifying AI agents. The framework is intended to establish requirements and accountability, while allowing payment networks to maintain their own verification and decision making processes.
The work will build on systems already developed separately by the three companies.
Visa has developed its Trusted Agent Protocol to help merchants identify trusted AI agents and support secure transactions involving them.
Mastercard has developed Verifiable Intent, which is designed to provide evidence of the user's authorised intent when an AI agent carries out a transaction.
Ant International has developed its Agentic Mobile Protocol, which is designed to support AI agent transactions through digital wallets, banking applications and other mobile platforms.
The new collaboration aims to develop common principles that can work across these different systems. The companies have not announced a completed global standard, and the framework remains at the development stage.
Ant International, Mastercard and Visa will work through BuildFin.ai, a platform convened by the Monetary Authority of Singapore. The platform brings together financial institutions, technology providers and researchers to develop AI solutions for financial services.
Singapore has also been working on safeguards for AI agents in financial services.
On July 3, the Monetary Authority of Singapore and financial industry participants published an industry white paper called Safeguards for Agentic Finance at Runtime, or SAFR. The framework proposes safeguards to help AI agents carry out financial tasks within predefined mandates, policies and risk limits.
The SAFR framework is an industry reference approach and is not regulatory guidance or a binding rule.
The growth of AI agents is creating new questions for the payments industry. In a conventional online purchase, a person normally selects a product, checks the details and authorises the payment. With an AI agent, some or all of these steps could be carried out automatically.
This creates questions about identity, permission and responsibility. Payment companies need to establish whether an agent is authorised to act for a customer, what the customer has allowed it to do and whether a transaction remains within those instructions.
The proposed KYA framework is intended to address part of this challenge by developing common approaches for identifying agents and assessing their trustworthiness.
The companies cited a McKinsey estimate that AI agents could potentially orchestrate between $3 trillion and $5 trillion in global consumer commerce by 2030. The figure is a projection, not an established market value.
Several questions remain unresolved, including how widely the proposed framework will be adopted, how AI agents will be certified and how responsibility will be assigned when an agent makes an unauthorised or unexpected purchase.
For now, the announcement marks an effort by two major payment networks and Ant International to address these issues as AI agents become a larger part of digital commerce.