Menu

Mastercard Enhances Agent Pay Framework to Secure AI-Driven Transactions Across Global Commerce

Terry KS 22 hours ago
Mastercard has announced an expansion of its Agent Pay trust framework, introducing new intelligence and risk verification features to provide clearer context for AI-initiated transactions. Designed to reduce friction and prevent fraud, these tools allow financial institutions and merchants to evaluate transaction intent, behavior, and risk as agent-led commerce accelerates globally

MALAYSIA, 12 OCTOBER 2026 – Mastercard has officially unveiled the expansion of Agent Pay, its agentic payments initiative, introducing advanced trust and intelligence capabilities created to establish greater context for transactions initiated by artificial intelligence. By integrating identity, behavioral analysis, intent, and fraud detection signals into a unified system, the technology equips merchants and financial institutions with deeper insights to support accurate authorization decisions and minimize unnecessary transaction friction.

The newly deployed services constitute the core intelligence layer of the Mastercard Agent Pay Trust Framework, an enterprise infrastructure unifying controls, identity, execution, intent, and risk intelligence. The rollout arrives as automated commercial interactions gain significant traction, with projections indicating that 10 percent of consumers worldwide will routinely rely on AI agents to execute purchases by 2030.

To address the complexities of agentic shopping, Mastercard is piloting a proprietary probability scoring model in the United States. The probability score assesses the likelihood that a payment was triggered by an AI assistant, granting issuers greater certainty to approve legitimate automated purchases. Over time, the model will incorporate broader datasets, including consumer propensity, credential risk, merchant risk metrics, and behavioral patterns across the full transaction lifecycle.

The shared intelligence framework actively answers critical verification questions during authorization, including whether transaction details match expected behavior, if any unusual markers exist on the agent or merchant side, and whether additional authentication steps are necessary. For instance, if an AI agent schedules an entire travel itinerary consisting of separate flight, hotel, and transport bookings simultaneously, Mastercard provides the contextual risk assessment needed for banks to approve the sequence without triggering false fraud declines.

To ensure scalable industry deployment, Mastercard is collaborating with web security provider Cloudflare to develop privacy-preserving analytical environments combining payment network data with web traffic signals. Additionally, Mastercard is partnering with Skyfire, a specialist in Know Your Agent (KYA) identification technology, to help financial networks verify AI agent identities, ensure accountability, and streamline authorization across the digital ecosystem.

%d