Mastercard spent decades training its fraud system to see bots as thieves. Now bots are the ones doing the buying.

Mastercard is flipping its entire fraud playbook. For decades, they've fought bots, but now they're building the infrastructure to let AI agents actually spend money.

Agentic commerce is moving from cool demos to the real economy. If agents can't prove intent and identity to banks, the entire autonomous economy stalls before it even starts.

This shifts the battlefield from model quality to transaction trust. Established players like Mastercard and Visa are racing to define the 'Know Your Agent' (KYA) standard.

• →Building the verification layer for agents, specifically tools that provide cryptographic proof of intent.
• →Developing B2B procurement agents that handle inventory and replenishment without human intervention.
• →Creating security tooling that automates the detection of model drift in financial agents to prevent rogue spending.

• →The 'Wrong-Nike' problem, where agents make irreversible, non-returnable purchases that violate user intent.
• →New attack vectors where sophisticated actors use high-fidelity agents to exploit the new KYA protocols.