The Linux Foundation says it intends to launch a new open standard called the Agent Name Service, or ANS, aimed at giving AI agents a trusted identity layer as they spread across the internet.
The group announced the plan on Monday, saying ANS would extend the existing Domain Name System, better known as DNS, to help users and systems verify and discover autonomous software agents. The foundation framed the project as a way to address identity, authentication, and governance challenges before agentic AI becomes more deeply embedded in enterprise and consumer systems.
Rather than creating a separate registry, ANS is designed to sit on top of DNS, the internet’s long-standing naming infrastructure. The Linux Foundation said the approach would avoid centralized control and proprietary naming systems while allowing organizations to identify agents through domain names they already own.
According to the announcement, the framework is intended to let operators check who an agent represents, what permissions it has, and whether its code and operational history have stayed intact. The organization said that kind of verification is increasingly important as AI agents move beyond experiments and into production use.
Jim Zemlin, chief executive of the Linux Foundation, said trusted identity infrastructure is becoming a basic requirement as AI agents begin operating across enterprises, platforms, and digital services. He said ANS is meant to provide a scalable and interoperable framework by relying on open standards.
The foundation pointed to industry interest in AI agents, citing World Economic Forum data saying 82% of executives plan to adopt them within the next one to three years. It said that level of interest has outpaced the development of secure methods for evaluating and managing autonomous systems.
The Linux Foundation said ANS is meant to be a federated system that supports identity and discovery without requiring a new proprietary ecosystem. The project also supports decentralized identifiers, or DIDs, and Legal Entity Identifiers, or LEIs, so organizations can connect existing identity tools into one verification model.
Supporters of the project argued that open internet standards are the right foundation for the next phase of AI deployment. Jared Sine, chief strategy and legal officer at GoDaddy, said the internet’s growth depended on open standards and shared infrastructure, not closed systems. He said ANS could help agents be identified and discovered across the open web.
Other supporters said the effort could reduce what they described as shadow AI risks and make it easier for enterprises to adopt agents without building separate integrations for each vendor. Cloudflare, Cisco, Salesforce, Infoblox, Hashgraph Online, and DistributedApps.ai were among the organizations whose representatives offered supporting comments in the announcement.
The Linux Foundation said the project is looking for participation from enterprises, AI developers, infrastructure providers, and security researchers. Technical repositories and contribution details are expected to be shared through the Agent Name Service GitHub organization.
The announcement did not provide a launch date for ANS, but positioned the project as an early effort to establish identity infrastructure for what it called the agentic web. The foundation said the goal is to keep that infrastructure open, neutral, and compatible with the systems the internet already uses.