As AI agents begin to act across websites, apps, and business systems, every interaction needs a simple way to confirm which agents are legitimate and who operates it. ANS is an open standard that assigns agents identity through domain names and DNS.
Your agent books a barber before your next haircut. A finance agent pays an approved supplier. A procurement agent compares vendors. A customer support agent resolves an issue with another company's agent.
This is where AI is headed: agents moving across services, systems, and organizations to get work done.
Identities are foundational to how the web works. Websites have domains. People have accounts and profiles. Businesses have verified digital properties. Without clear identities, it's hard to verify who or what you're interacting with.
And right now, agents have nothing standard. As more come online, there is no shared way to identify them across websites, apps, or platforms.
Without a standard, every agent interaction needs to ask the same questions, each time: Is this agent what it claims to be? How should it be handled?
Every actor on the web has an identity standard. Agents do not — yet.
ANS is an open standard that connects an AI agent to a domain name, creating a recognizable and verifiable identity that can be discovered through DNS.
That means an agent can show who or what it represents, where its identity is anchored, and how other agents, platforms, and systems should interact with it.
Instead of every company inventing its own way to identify agents, ANS gives the ecosystem a shared starting point.
Discovery and trust are different problems. ANS is the trust layer that works with any discovery mechanism.
Infoblox (DNS-AID), GoDaddy (ANS), and Nemethi (AID) are actively collaborating on a shared DNS profile under the _ag label. This is the first concrete evidence that the fragmented landscape is consolidating rather than continuing to diverge. The convergence on shared SVCB records — one record per protocol, carrying discovery metadata for all participating drafts — provides a strong trust signal for anyone evaluating whether to build on ANS.
Just like websites use domain names and DNS to make it easy for people to find them, ANS extends that familiar model to AI agents, using domain names and DNS as the foundation for agent identity. And the infrastructure is already built and being used.
The idea is simple: if an agent acts for an organization or individual, its identity should be anchored to a domain.
Help shape how AI agents identify themselves on the internet. Get in touch or explore how ANS works.