THE ESSENTIALS IN 30 SECONDS

  • The Linux Foundation announced Agent Name Service (ANS) as a hosted project on November 10, 2026, at KubeCon North America.
  • The specification spans seven documents, from domain-anchored naming to agent-to-agent authentication.
  • Eleven organisations, including Cisco, Cloudflare, DigiCert, GoDaddy, Salesforce and OWASP, have contributed to or adopted it.

The Linux Foundation has formally launched Agent Name Service (ANS) as one of its hosted projects, aiming to give AI agents identities that other systems can verify. The announcement was made on November 10, 2026, at KubeCon + CloudNativeCon North America in Salt Lake City, five months after the foundation first said it intended to take the project on.

How ANS works

Rather than creating a new naming system, ANS builds on the Domain Name System. An agent's identity is anchored to a domain, published through DNS and backed by the same kind of certificates that websites already use to prove who they are. The specification is split into seven documents covering domain-anchored naming, registration and versioning, DNS publication, auditing through public logs, integrity monitoring to detect unauthorised changes to identity records, and authentication between agents.

The problem it targets is straightforward: before an organisation lets an autonomous agent touch its systems or data, it needs to know which agent it is dealing with and whom that agent represents. When the intent to launch was announced in June, the foundation said ANS would also support decentralised identifiers and Legal Entity Identifiers, and would avoid centralised or proprietary registries.

Who is behind it

Eleven organisations are listed as having contributed to or adopted the standard: Cisco, Cloudflare, DigiCert, DistributedApps.ai, GoDaddy, HOL, Infoblox, Radix, Salesforce, OWASP and Trustmodel.ai. Ken Huang of DistributedApps.ai, lead author of the research paper that introduced ANS, led its initial development and plans to join the technical steering committee, which oversees the specification as it is developed in the open. A DigiCert executive described the system as "a passport for AI agents".

Early integrations exist. AGNTCY, another Linux Foundation project, supports ANS identity checks in its Agent Directory Service, and a tool called ANS Finder implements the Agentic Resource Discovery specification and passes its registry conformance tests. Trustmodel.ai says it is tying its agent trust scores to ANS identities.

What remains unclear

The announcement gives no version number, maturity level or timeline for the specification, and no figures on production deployments. ANS addresses identity and discovery; deciding what a verified agent is allowed to do is left to other systems. Its relationship with agent protocols such as MCP or Agent2Agent was not described. As with earlier attempts to standardise online identity, adoption by agent platforms and enterprises will determine whether a shared DNS-based approach takes hold or vendor-specific schemes continue to multiply.