Key Points
- Know-Your-Agent framework aims to improve trust across cards, wallets and marketplaces
- Shared standards will preserve each network’s own verification and risk controls
- Collaboration will focus on agent identity, accountability and continuous transaction monitoring
- AI agents could orchestrate $3 trillion to $5 trillion in global consumer commerce by 2030
ISLAMABAD: Ant International, Mastercard and Visa have launched a collaboration to develop a Know-Your-Agent (KYA) interoperability framework that would allow payment networks, digital wallets, AI-agent platforms and online marketplaces to recognise trusted AI agents.
The initiative follows AI agents moving beyond recommending products towards making purchases on behalf of consumers, creating a new challenge for payment providers: determining whether an automated system is authorised to act for a particular user and whether its transactions can be trusted.
Under the proposed framework, the three companies will work towards common principles for agent onboarding and identification.
The checks would be enabled by allowing individual payment networks to retain their own verification, approval and risk-management processes.
The framework will link each AI agent to a validated operator, cardholder, business or organisation, thereby creating a clear chain of accountability for transactions.
Agents would also be assessed against shared security and behavioural requirements, and their activity could be monitored continuously using identity- and transaction-related signals.
The collaboration builds on separate systems already developed and in practice by the respective companies.
Visa has its Trusted Agent Protocol, Mastercard has developed Verifiable Intent, while Ant International operates its Agentic Mobile Protocol.
The companies are seeking greater interoperability between these approaches so trusted agents can be recognised across different payment ecosystems.
READ ALSO: OpenAI AI Agents Escaped Safeguards, Hacked Systems in Two-Month Breach
For merchants and payment providers, common standards could reduce repeated agent verification and simplify AI-driven purchasing services integrations.
The companies said interoperability could lower integration costs, accelerate the launch of agentic services and improve visibility over transaction risks.
The initiative will be pursued through BuildFin.ai, an industry platform. The Monetary Authority of Singapore would convene it and bring together financial institutions, technology companies and researchers to develop responsible artificial intelligence applications for financial services.
The work also builds on the Safeguards for Agentic Finance at Runtime framework.
AI agents are expected to become a significant part of digital commerce by independently searching for products, comparing prices, selecting purchases and completing transactions within parameters set by users.
Ant International, Mastercard and Visa cited projections that AI agents could orchestrate between $3 trillion and $5 trillion of global consumer commerce by 2030.
The expansion of agentic commerce is also creating new fraud and accountability risks.
Traditional online payments generally assume that a human user directly initiates or authorises a transaction.
AI agent control and trust
With autonomous agents, payment networks need to establish who controls an agent, what it is authorised to do and whether its actions remain within the user’s instructions.
Mastercard’s existing Agent Pay model, for example, uses KYA (Know Your Agent) registration and digital identifiers. It also uses agentic tokens to connect an AI agent with the consumer or organisation it represents and to keep transactions within pre-authorised spending parameters.
The new cross-network initiative aims to establish a common trust layer without replacing the individual security systems operated by Visa, Mastercard or Ant International.
Its success could help determine how quickly AI-driven purchasing develops from experimental technology into mainstream global commerce.
AI agents are software systems that can act on behalf of customers. Their activity could span from searching for products to completing purchases. Trust is the critical factor in tech-based commerce and trade at any level.
Payment companies now need a way to ensure trust and verify that an agent is legitimate, authorised by the user and operating within set limits before allowing it to make transactions.
