During the second quarter of 2026, a prominent cybersecurity network recorded an impressive 17.7B requests driven by AI-agent traffic. This volume represents a substantial 45% increase compared to the previous quarter. As these automated assistants become more prevalent, it is increasingly clear that each agent requires a distinct identity coupled with transparent, accountable authorization from its human user. Yet, the industry currently lacks a reliable method to securely link a digital agent directly to the individual it acts for and the established credentials of that person.

Several initiatives are presently tackling distinct components of trust within these automated interactions. For instance, Visa TAP focuses on authenticating agents and confirming their related authorizations. Meanwhile, Google AP2 relies on signed mandates to accurately record the intentions of human users. OpenAI ACP manages the delegation of both payments and authentication. In addition, IETF Web Bot Auth provides authentication for automated clients, though it explicitly excludes the authentication of the end user.

To bridge these existing gaps, AIR is developing a unified framework. Under this comprehensive system, an agent receives a unique identity that links securely to the AIR Identity of the user. Individuals grant a revocable delegation that clearly outlines the specific merchant, time window, limits, and allowable scope for the agent. Consequently, the automated assistant only shares the exact information permitted by this delegation, ensuring the complete identity record of the individual remains private. This approach significantly minimizes data custody risks, as applications can validate the credential, the delegation, and the entire revocation chain without ever needing to consult a centralized identity directory.

We invite you to reach out and schedule a demo by visiting https://air3.com/partner-with-us