Meta and partners launch an open standard for AI agent authentication

Meta is launching an open standard for AI agent authentication together with a group of partners, according to a report by Quartz.

The wording — “launching” — describes a multi-party standards effort led by Meta rather than a single product release.

The focus is authentication for AI agents. As agents begin to act on behalf of users and interact with online services, reliably establishing an agent's identity and permissions becomes a pressing problem.

Choosing an open standard rather than a proprietary protocol suggests Meta wants the mechanism adopted across the industry, not confined to its own ecosystem. For platform companies, backing an open standard is also a way to expand influence over how the technology is governed.

The move also reflects a broader shift in attention: from raw model capability toward the trust and security layer that agents need to operate at scale. As agents touch payments, commerce, and enterprise systems, that identity layer becomes more important.

What to watch next: which partners are involved, how the standard is specified technically, and whether other major AI vendors and platforms adopt it.