Visa, Mastercard, and Ant International are collaborating on a Know-Your-Agent interoperability framework designed to help card networks, digital wallets, and agent platforms verify trusted AI agents across various systems without compromising their own risk controls.
The framework will reference Visa's Trusted Agent Protocol, Mastercard's Verifiable Intent, and Ant International's Agentic Mobile Protocol. According to the companies, shared standards could reduce integration costs and enable more effective risk management. The framework is being developed through BuildFin.ai, a platform launched by the Monetary Authority of Singapore.
Trust and Liability Questions
A joint report from Visa and Artemis indicates that payment infrastructure alone is insufficient to address the challenges posed by autonomous agents. Traditional commerce assumes a person is responsible for each transaction, but AI agents blur this responsibility. If an agent is hacked or misdirected, liability could fall on the user, platform, model provider, or merchant.
The US National Institute of Standards and Technology is also examining the issue, publishing a concept paper in February that recommended better controls for software agents regarding identification, authorization, auditing, and accountability, including defense against prompt-injection attacks.
Cryptocurrency and Agent Payments
Cryptocurrency networks could become significant venues for agent transactions. According to Chainalysis, x402 payments on Base exceeded 100 million transactions in three quarters, though most were attributed to memecoin farming rather than sustained commercial demand. TRM Labs reported global retail crypto activity at approximately $979 billion in Q1 2026.
The Bank for International Settlements explored AI's potential in a working paper, demonstrating that general-purpose AI could perform high-level intra-day liquidity management in wholesale payment operations, while cautioning that reliability, accountability, human supervision, and cyber resilience remain critical.
Safety and Regulatory Considerations
On-chain transactions are more difficult to reverse than card payments, raising stakes if agents malfunction or are compromised. OpenAI has called for legislative regulation based on AI capabilities, including independent evaluation, cybersecurity measures, and incident reporting.
Consumer confidence remains limited. A Cryptopolitan survey found that 30.22% of respondents would not allow AI to manage their wallets, while 25.9% would permit it from a reliable company. Stanford University's 2026 Artificial Intelligence Index reported that only 31% of US respondents believe their government can regulate AI responsibly, compared to a global average of 54%.
