EMVCo Publishes Draft Framework for AI Agent Card Payments, Seeks Industry Feedback

The Quick Take

• EMVCo releases draft framework for secure, interoperable card-based agentic payment systems
• The framework targets AI-agent-initiated payments, addressing authentication and authorization at machine speed
• Global card network standards body seeks industry input to shape the emerging agentic commerce infrastructure

EMVCo, the technical standards body that manages the EMV specifications underpinning virtually all global chip-and-PIN card transactions, has released a draft framework for agentic payments — transactions initiated autonomously by artificial intelligence agents on behalf of human principals without direct real-time human authentication at the payment step. The framework addresses a fundamental infrastructure gap created by the rapid growth of AI-agent commerce, where traditional card network security protocols designed for human-initiated transactions are ill-suited for machine-speed autonomous spending.

The initiative positions EMVCo — whose governance includes Visa, Mastercard, American Express, Discover, UnionPay, and JCB — to set the global standard for AI-agent payment security before proprietary systems fragment the market. Fintech companies building agentic commerce platforms, banks issuing AI-agent-linked credentials, and payment processors anticipating a new high-volume transaction category all have significant commercial stakes in the outcome. A unified framework would accelerate institutional adoption of agentic commerce infrastructure, benefiting companies with early payment-AI integration.

The key forward signal is the timeline from draft feedback to final specification — typically 12-18 months in EMVCo's process — and which major merchants and banks commit to early implementation. The macro variable determining whether agentic payments scale to meaningful transaction volumes within 2-3 years is the pace of enterprise AI agent deployment in procurement, subscription management, and expense categories where delegated machine spending creates clear efficiency gains over human-approved transactions.

Synthesized from 1 source.