The FIDO Alliance launches two working groups to establish industry standards for securing AI agent transactions; Google contributes the Agent Payments Protocol
AI agents may soon be buying your stuff for you. The FIDO Alliance has teamed up with Google and Mastercard to try to ensure …
Context & Ripple Effects
Google had already introduced the Agent Payments Protocol with support from more than 60 organizations, while broader agent standards work was also taking shape through efforts involving major AI and cloud companies. The FIDO Alliance’s working groups move the focus from agent-to-agent interoperability toward the security of transactions agents initiate.
The development also sits alongside partnerships such as OpenAI and Visa’s work on permissioned agent purchases. It matters because payment authorization and transaction security are becoming practical prerequisites for agents to act on users’ behalf.
First-order effects
- FIDO Alliance members, including Google and Mastercard, gain a forum to define common security approaches for AI-agent transactions; Google’s AP2 becomes an input to that process rather than a standalone protocol.
- Merchants, payment providers, and agent builders get a clearer industry venue for addressing how agent-led purchases are authorized and secured.
Second-order effects
- Competing agent-payment and commerce initiatives will face pressure to align with emerging FIDO-led requirements or demonstrate interoperability with them, particularly where they touch user permission and payment credentials.
- Security and identity layers become more central to agent-commerce deployments, shifting implementation attention beyond the model or assistant to transaction controls shared across platforms.
Third-order effects
- If common standards emerge, agentic commerce could develop around portable permission and authentication mechanisms rather than separate, platform-specific payment flows.
- The key structural question will be governance: standards shaped by major AI, payment, and identity participants may lower integration friction while concentrating influence over how agents are permitted to transact.
The trend: AI agents are moving from conversational interfaces toward permissioned economic actors, making interoperable identity, authorization, and payment-security standards a core layer of the market.