FIDO Alliance launches working groups for AI agent payment security standards

12 articles · Updated · WIRED · Apr 28

Google and Mastercard are contributing open source tools, including Google's AP2 protocol and Mastercard's Verifiable Intent framework, to accelerate the standards' development.
The initiative aims to create cryptographic and privacy-preserving mechanisms for authorizing, validating, and protecting AI agent-initiated transactions across industries, addressing risks like agent hijacking and unauthorized payments.
With rapid adoption of agentic AI, the FIDO Alliance, Google, and Mastercard stress the urgency of establishing interoperable protections to promote trust, transparency, and accountability in AI-powered commerce.

Could Google and Mastercard's new standards create a 'walled garden' for trusted AI commerce?

Beyond payments, how will these standards stop AI agents from being tricked into leaking sensitive personal data?

If an AI agent makes a rogue purchase, who ultimately pays for it under the new FIDO standards?

Will smaller businesses be able to afford the technology required to adopt these new AI security standards?

With AI evolving so rapidly, can any security standard truly keep pace, or is this a losing battle?

How do we legally define an AI's 'intent' when it acts autonomously on our behalf?