If you have ever used the internet, you have already relied on DNS without even realizing it
It is that silent system that translates an address like yoursite.com into an IP number that machines understand, working like a massive phone book for the web. You type a name, it returns an address. Simple as that, and so fundamental that practically the entire modern internet depends on it to function, even though the vast majority of people have never stopped to think about it for even a second.
Now imagine that AI agents are showing up in full force across the corporate world, making decisions, executing tasks, and communicating with other systems autonomously, and nobody knows for sure if they are who they say they are. 😬
That is exactly where the Agent Name Service, or ANS, comes in.
On June 23, 2026, the Linux Foundation announced its intention to launch this new open standard, built right on top of the existing DNS infrastructure, to solve a problem that is only going to get more urgent: giving verifiable identity to the AI agents operating across the internet.
This is not some futuristic solution waiting to happen. It is a direct response to a scenario that is already knocking on the door of businesses everywhere.
The problem nobody has really solved yet
When an AI agent sends a message to another system or takes an action on behalf of a company, how does anyone confirm that agent is legitimate? Today, the honest answer is: with a lot of difficulty. The rapid growth of multi-agent architectures, where multiple AI agents collaborate with each other to complete complex tasks, has created a massive trust vacuum. Until now, there has been no standardized, open, and scalable way to verify the identity of these agents, understand what they do, which protocols they communicate with, or how other systems can discover them in a secure and reliable way.
This vacuum is not just a technical inconvenience. It represents a real security risk for any organization adopting automation with AI agents. A malicious or compromised agent could impersonate a legitimate system, access resources, make decisions, and cause damage before anyone even noticed what was happening. And the more autonomous these agents become, the larger the attack surface available for exploitation, whether by external actors or internal configuration and governance failures.
According to data from the World Economic Forum, 82% of executives plan to adopt AI agents within the next one to three years, despite widespread uncertainties about how to evaluate and manage autonomous systems safely. That number shows the urgency of the problem: adoption is accelerating, but the trust infrastructure has not kept up with the pace.
The Linux Foundation identified this gap and decided to act with an approach that makes a lot of sense in practice: instead of inventing a completely new infrastructure, leverage what already works at a global scale. DNS is already resilient, already distributed, already widely adopted, and already has decades of accumulated security best practices. Building the Agent Name Service on that foundation means adoption can be much faster and the learning curve for technology teams is considerably lower than it would be with a solution built from scratch.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
How the Agent Name Service works in practice
The ANS works as an identity and discovery layer built directly on top of DNS, extending what the system already does very well — mapping names to resources — into a new context where the resources are autonomous AI agents. In practice, each AI agent registered on the Agent Name Service receives a unique and verifiable identifier, associated with metadata describing its capabilities, the communication protocols it supports, the access policies it enforces, and the credentials that prove its origin and authenticity. It is as if each agent gets an ID card recognized by every other system participating in the ecosystem.
The standard proposed by the Linux Foundation also includes mechanisms for agents to discover one another programmatically, without a human needing to broker every connection. This is essential in architectures where dozens or hundreds of AI agents need to coordinate in real time to complete complex workflows.
With ANS, an agent can query the service, verify another agent’s identity, confirm that it has the necessary permissions and capabilities, and only then establish communication — all in an automated and auditable way, which is exactly the kind of traceability that security and compliance teams need.
On top of that, the framework supports Decentralized Identifiers (DIDs) and Legal Entity Identifiers (LEIs), allowing organizations to integrate existing identity systems into a unified verification model. This matters because many companies have already invested in identity solutions and do not want to throw them away to adopt something new. The ANS fits into what already exists, rather than requiring everything to be rebuilt from the ground up.
One important detail is that the Linux Foundation proposal aligns with initiatives already gaining traction in the AI ecosystem, like Anthropic’s MCP protocol and Google’s A2A standard, both focused on agent-to-agent communication. The Agent Name Service does not compete with these efforts — it complements them, providing the identity and discovery layer that these communication protocols need to work reliably in real production environments, where trust cannot be assumed and must be verified in every interaction. 🔐
What major companies are saying about ANS
The Agent Name Service announcement did not come alone. Several heavyweight technology companies have publicly voiced their support for the project, which gives a pretty good sense of the level of interest this initiative is generating.
Jim Zemlin, CEO of the Linux Foundation, emphasized that AI agents will increasingly operate across companies, platforms, and digital services, making trustworthy identity infrastructure a fundamental requirement. According to him, by building on DNS and open standards, ANS creates a scalable and interoperable framework for verified agent communication in the global digital economy.
Cloudflare, through its CTO Dane Knecht, reinforced its commitment to supporting the open standard to ensure that the agentic web is fast, secure, and built on a shared foundation of trust. Knecht pointed out that the global success of the internet is rooted in open, shared principles, and that the next phase, powered by autonomous agents, needs to follow the same path.
Jared Sine, Chief Strategy and Legal Officer at GoDaddy, brought an interesting perspective by stating that the success of the internet did not come from proprietary systems, but from open standards, shared infrastructure, and an ecosystem committed to working together. In his view, ANS creates a path for agents to be identified and discovered on the open web, helping ensure that the next era of innovation remains as open and interoperable as the last one.
Salesforce also weighed in. Srini Tallapragada, President and Chief Engineering & Customer Success Officer at the company, got straight to the point: identity is imperative for enabling AI agents to operate on the open web, and ANS defines a common standard for identity and verification that lets developers move faster with trust, interoperability, and security built in from the start.
Nathan Jokel, Senior Vice President of Corporate Strategy at Cisco, put his finger on a central issue. According to him, every platform shift creates a choice between walled gardens and open ecosystems, and agentic AI is no exception. Cisco is contributing to open standards efforts at the IETF and the Linux Foundation because the value of this technology only multiplies when the ecosystem is open, decentralized, and secure.
Ken Huang, CEO of DistributedApps.ai and lead author of the award-winning research paper on the Agent Name Service, shared that his biggest concern was that agentic AI systems would proliferate without a neutral, trusted layer for identity and discovery, creating exactly the shadow AI risks that keep security teams up at night. In his view, the Linux Foundation governance model is precisely what ANS needs.
Why the Linux Foundation is at the center of all this
The choice of the Linux Foundation as the steward of this standard is no accident. The organization has a solid track record of creating and maintaining open infrastructures that have become pillars of modern technology, from Linux itself to Kubernetes, along with projects like OpenSSF, PyTorch, RISC-V, and SPDX. When the Linux Foundation embraces a standard, it brings along a global community of developers, companies, and organizations that trust the neutral governance and commitment to openness that the institution represents.
This is especially important for an identity standard, where the neutrality of the maintainer is just as crucial as the technical quality of the solution. Nobody wants a single company controlling who can register AI agents, which credentials are accepted, or how trust policies are defined. An identity standard needs to be governed neutrally so that all ecosystem participants feel comfortable adopting it, and the Linux Foundation has decades of experience doing exactly that.
Additionally, the fact that ANS is an open standard means that any company, regardless of size or industry, can implement it, contribute to it, and audit it. There is no vendor lock-in, no single company controlling who can or cannot register their AI agents on the service, and no black box managing critical identity and trust decisions. This openness is what will largely determine whether the standard truly takes off or remains confined to an early adoption niche in the coming months.
The timing is also quite strategic. The AI agents market is at a stage where companies are still defining their architectures, frameworks, and governance policies for autonomous AI. Establishing a standard now, before fragmented proprietary solutions become entrenched, is the window of opportunity the Linux Foundation is seizing. If the Agent Name Service gains enough adoption over the next 12 to 18 months, it has the potential to become an infrastructure as fundamental to the AI ecosystem as the original DNS is to the internet — something that just exists, everyone uses, and everyone trusts. 🚀
The role of DNS as a trust anchor
One of the smartest aspects of the ANS approach is the decision to anchor agent identity directly in DNS, rather than creating a completely separate registration system. DNS already processes more than 100 million queries per second worldwide, and its globally distributed infrastructure is one of the most robust and battle-tested in the history of computing. Leveraging that foundation is not just a technical convenience — it is an architectural decision that immediately solves several scale, availability, and resilience problems that any new identity system would face if built from scratch.
As Wei Chen, Chief Legal Officer and Executive Vice President of Regulatory Strategy at Infoblox, pointed out, ANS allows any organization to identify its agents through the domain names it already owns, with no proprietary namespace and no gatekeeper. DNS becomes a trust signal for agents, anchored in the same proven mechanisms that already protect the web. A shared specification, built on DNS, open to every organization and subordinate to none.
This approach also makes corporate adoption much easier. Infrastructure teams already understand DNS, already know how to manage records, and already have integrated monitoring and security tools. Adding AI agent identity records to that existing infrastructure is a natural extension, not an operational revolution that requires extensive training or radical changes to network architecture.
What this means for anyone building with AI right now
For engineering and architecture teams already developing solutions with AI agents, ANS represents an important shift in perspective. Until now, agent identity has been handled in an ad hoc fashion, with each team solving the problem their own way, using tokens, API keys, or proprietary mechanisms from their chosen infrastructure providers. That fragmentation does not scale and creates silos that make interoperability between different systems harder, especially in scenarios where agents from different vendors need to collaborate within the same corporate workflow.
With an open standard like the Agent Name Service, teams get a common reference for implementing verifiable identity in their agents, which makes not just system-to-system communication easier, but also auditing, monitoring, and incident response. When something goes wrong in a multi-agent system — and eventually something always goes wrong — having a clear record of which agents were operating, with what permissions, and on whose behalf is the difference between an investigation that takes hours and one that takes weeks.
The project is already seeking participation from companies, AI developers, infrastructure providers, and security researchers interested in helping establish open standards for the emerging agentic ecosystem. Technical repositories and contribution opportunities are available through the Agent Name Service organization on GitHub.
The next steps for the agentic ecosystem
The launch of ANS is a milestone, but it is only the beginning of a longer journey. The real test will come in the months ahead, when the standard needs to prove it works in real production environments, under heavy workloads, adversarial scenarios, and sophisticated attacks. The theory behind the Agent Name Service is solid, but practical implementation at scale is where the hard details show up.
The names that have already publicly positioned themselves alongside the project — Cloudflare, GoDaddy, Salesforce, Cisco, Infoblox, and Hashgraph Online — suggest there is enough critical mass to at least guarantee a robust initial ecosystem. But the real test will be adoption by smaller companies, startups, and independent developers, because a standard that only works for large corporations is not really a universal standard.
As Michael Kantor, President of Hashgraph Online, summed it up well, the next phase of the internet will depend on shared standards that allow agents to move between protocols and ecosystems without creating new silos. ANS is a significant step in that direction.
It is worth watching closely how the project evolves within the Linux Foundation and which organizations publicly commit to adopting the standard in the coming weeks. The strength of an open standard is measured by the community that embraces it, and the first names that join the effort will give a solid indication of whether the Agent Name Service is on track to become a real cornerstone of AI agent infrastructure or just another well-intentioned standard that could not build enough critical mass to change the game. 👀
