--- title: "The Internet of Agents. Who Gets to Name Them?" date: 2026-11-12 author: "Erwin Groen" description: "As AI agents begin acting on our behalf, naming, identity and accountability are becoming increasingly important. Could Handshake offer an independent alternative?" image: "/media/internet-of-agents-accountability.png" tags: - Handshake - ArtificialIntelligence - AgenticAI - DigitalIdentity - HNS --- *As artificial intelligence begins doing more things for us, knowing who or what we're dealing with is becoming very important.* *🎬 Tendo introduces the questions of Agentic Identity, Handshake and accountability.* The internet is evolving. For years, we've used it to visit websites, send emails, buy things and communicate with one another. Artificial Intelligence is beginning to do some of those things for us. We're moving towards a world where AI agents can book appointments, make purchases, carry out business tasks and communicate with other AI agents, sometimes without us needing to guide every step. This is termed: **The** **Internet of Agents**. It sounds convenient. And in many ways, it will be. But it raises a question.. **Who gets to name these agents?** ## Who Are We Dealing With? Imagine you're driving along a road and ask an AI agent whether it's safe to travel at a certain speed. The agent tells you it is. You trust it, but it's wrong. The road conditions make that speed dangerous, an accident follows. **Who's responsible?** ⚪️ You, for trusting the advice? ⚪️ The company that developed the AI agent? ⚪️ Or the person or organisation operating it? How would you even know who was behind the agent in the first place? These questions matter because AI agents are becoming capable of doing more than answering questions. They can influence decisions and increasingly take action on our behalf. Before we place our trust in them, we need ways to identify who operates them and establish who can be held accountable when things go wrong. That starts with *identity*. A name alone won't establish legal responsibility, but it can be a starting point for finding out who's behind an agent. ## Who Controls Their Names? Today, most of us use internet names and services that depend on companies or organisations to keep them working. Your website has a domain name. Your email has an address. Your social media account has a username. In many cases, you don't have complete control over those names. There are rules, renewal fees and organisations that can decide whether you are allowed to keep using them. Imagine a future where millions of AI agents need names of their own. ❓Will they all need to register through a handful of large companies? ❓Will one organisation decide which names can exist? ❓And what happens if the rules change? ⚪️ We could end up creating another system where someone else controls an important part of our digital identity. ## What Does Handshake Have to Do With This? This is where [Handshake](https://handshake.org/) becomes part of the bigger picture. Handshake is an alternative internet naming system. Instead of relying on the existing system managed by ICANN to approve and introduce new top level names, Handshake allows people to obtain and control names through a public blockchain, subject to the protocol's rules. Think of it as another way of establishing control over an internet name, without having to ask one central organisation to issue it. For AI agents, this could be useful. An individual, business or organisation could use a Handshake name as part of the identity of an AI agent operating on its behalf. Of course, simply having a name doesn't prove an agent is genuine. Just as someone can create a website with an impressive name and still be dishonest, an AI agent would need additional ways to prove who operates it and what it is allowed to do. But having greater independence over the name itself is a useful starting point. Handshake also has practical limitations. Its names aren't automatically recognised by every browser or internet service, and more work would be needed to make them useful across AI systems. Still, the opportunity is there. ## Do We Really Need Another Gatekeeper? New companies and organisations are already working on ways to identify and manage AI agents. Some of these systems may become very useful, especially when it comes to security and protecting people from fraud. But there is a difference between making sure an agent is trustworthy and allowing one company to decide who can have a name. We should be able to check an agent's identity without necessarily giving a single organisation control over that identity. That's an important distinction. Because once a particular system becomes widely accepted, changing to something else can become difficult. We've already seen how dependent people can become on social media platforms, domain registrars and other online services. Do we want the same thing happening with AI agents? ## The World Is Already Asking These Questions And this isn't just speculation about what might happen in the future. On 10 November 2026, the Linux Foundation [announced](https://www.linuxfoundation.org/press/linux-foundation-welcomes-agent-name-service-to-bring-verifiable-identity-and-discovery-to-ai-agents) that a project called Agent Name Service had joined its organisation. Its purpose is to help AI agents establish their identities, recognise one another and provide ways to verify who they represent. The system uses existing internet domain names as part of that process. Importantly, it is designed as an open approach rather than a single company's proprietary register. It's an important development because it demonstrates something we've been discussing at Own The Dot. **Naming could become an essential part of how AI agents establish trust.** The US National Institute of Standards and Technology has also been [examining](https://www.nist.gov/blogs/cybersecurity-insights/back-future-why-agentic-ai-needs-strong-identity-foundation) how AI agents should identify themselves and what they should be allowed to do. The [International Telecommunication Union](https://www.itu.int/en/mediacentre/Pages/PR-2026-07-09-focus-group-agentic-AI.aspx), a United Nations agency, has also begun work on trust and identity for AI agents. These organisations are approaching the problem in different ways, they're recognising the same challenge. ❓As AI agents become more independent, how do we know who we're dealing with? ❓Who gets to control the names they use? There is no confirmed integration between Handshake and Agent Name Service. But Handshake's independent approach to naming deserves consideration alongside these emerging systems. Because while others are developing ways to verify an agent's identity, we should also be asking who controls the name that identity depends on. ## An Opportunity Before the Rules Are Written Tendo, who kindly recorded the introduction to this article, raises an important point. As AI agents become more capable, knowing who or what we're interacting with will become increasingly important. And so will knowing who is responsible for their actions. Handshake doesn't provide all the answers. But it does offer something worth considering before the Internet of Agents becomes firmly established. A different way to own and control an internet name. The decisions being made today could influence how digital identities are managed for years to come. ## A Personal Thought I believe the Handshake naming protocol offers something increasingly valuable in today's digital world. An opportunity to obtain and control an independent internet name, without relying on a central authority to approve its existence or continued ownership. Handshake was designed to operate without a single organisation or government controlling its naming system. That independence makes it more resistant to outside interference and decisions made by those in positions of power. As artificial intelligence develops and new rules surrounding digital identity begin to emerge, I believe that independence deserves attention. We have an opportunity to consider how names should be owned before someone else decides how they should be issued. That's what makes Handshake important to me. **A unique opportunity to own an independent name. A chance to *Own your Own Dot*** That's the thinking behind **OWN THE DOT™**. The Internet of Agents is coming. **Before someone else decides how every agent should be named, we should at least explore the alternatives..** ## Further Reading - [Agent Name Service — Linux Foundation](https://www.linuxfoundation.org/press/linux-foundation-welcomes-agent-name-service-to-bring-verifiable-identity-and-discovery-to-ai-agents) - [AI agent identity — NIST](https://www.nist.gov/blogs/cybersecurity-insights/back-future-why-agentic-ai-needs-strong-identity-foundation) - [Handshake Protocol](https://handshake.org/) ## Tendo at Freedom Fest 2026 Tendo kindly contributed the video introduction to this article and is scheduled to speak at Freedom Fest in Manchester, United Kingdom, on 21 November 2026. His presentation, *“Handshake: The Foundation Of A Decentralized Internet,”* will explore the potential of Handshake as an alternative internet naming protocol. Freedom Fest brings together people interested in privacy, digital independence and technologies that give individuals greater control over their digital lives. **Event:** [Freedom Fest 2026](https://freedomfest.now/) **Presentation:** Saturday, 21 November 2026, at 2:50 pm UK time. Programme times are subject to change. [View the official event programme](https://bitfest.uk/programme/) --- *Own The Dot. Your name. Your choice. Your digital independence.*