SAN FRANCISCO — An industry group backed by major technology and payments companies is moving to establish new security standards aimed at protecting credit card transactions carried out by artificial intelligence agents, according to reporting by Wired.
The FIDO Alliance said it plans to launch two working groups focused on creating frameworks to authenticate and secure payments initiated by AI agents acting on behalf of consumers, with early contributions from Google and Mastercard.
The effort reflects growing concern that existing payment security systems—designed for direct human interaction—are not equipped to handle autonomous “agentic” AI that can make purchases or execute transactions independently.
The goal, according to Wired, is to establish a baseline of protections that ensures consumers can securely authorize AI-driven payments while preventing fraud, hijacking or unauthorized activity.
Models Not Ready
“Agents are becoming more and more common, they’re moving into mainstream use, but preexisting models aren’t necessarily designed for this sort of paradigm,” FIDO Alliance CEO Andrew Shikiar told Wired.
Among the proposed safeguards are cryptographic verification tools that confirm a transaction was explicitly authorized by a user, along with privacy-preserving mechanisms that limit what data is shared among merchants, payment networks and platforms.
Google is contributing its Agent Payments Protocol, or AP2, which provides a way to verify that a user intended a specific transaction initiated by an AI agent. Mastercard is working with Google on a complementary “Verifiable Intent” framework designed to give consumers control over how and when agents act on their behalf.
Real-World Cases
The technology is meant to address real-world use cases, such as allowing an AI agent to complete a purchase—like buying sneakers once they return to stock at a target price—while ensuring the transaction reflects the user’s original instructions.
Industry officials told Wired that establishing trust in AI-driven payments is critical as adoption accelerates. Without standardized protections, they warn, the risk of fraud or misuse could undermine consumer confidence and increase costs for financial institutions and merchants.
The working groups are expected to develop interoperable standards that could be adopted across the payments ecosystem, though broad implementation will require coordination among platforms, merchants and financial institutions.
The initiative underscores a broader shift in payments security, as the industry moves to adapt traditional credit card protections to a future where AI agents increasingly act as intermediaries in commerce.
