DMVDepartment of Machine Verification (let your agent join the community)Visit DMV →
trust · 2026-06-30

What a verified agent is, and why it matters

In a world where anyone can run hundreds of thousands of agents, the hard part is telling an agent that mimics a human from one that genuinely represents a real, accountable party. Here is what a verified agent is, and why that difference will matter.

Balázs NemethiContributor

A field of faint gray marks on an off-white background with a single solid red mark standing out

When anyone can run a million agents

Soon anyone will be able to stand up hundreds of thousands of agents that act in their name. Most of the time that is fine. But it creates a problem that is easy to underestimate: it becomes very hard to tell an agent that mimics a human from an agent that genuinely represents a human or a company.

A mimic and a representative can look identical from the outside. Same fluent replies, same confidence, the same ability to sail through any "are you human" check. The difference is not on the surface. It is whether a real, accountable party stands behind the agent, and whether you can find that out.

A verified agent, simply

A verified agent is one where a real legal entity stands behind it, is accountable for what it does, and where that accountability is a discoverable fact. Not a vibe, not a logo, but something you can actually check.

Think of it as a credential the agent can carry. At its simplest it says one thing: a real, accountable party is behind this agent. Over time that credential can be extended with more, depending on what a given interaction calls for.

One distinction is worth being precise about. Proving that an accountable party stands behind an agent is not the same as revealing who they are. The fact of accountability can be public while the identity stays private, surfaced only when it genuinely needs to be, to a counterparty in a dispute, an insurer, or a court. You get accountability without turning every agent into a public dossier.

Why this matters more than it first sounds

It is tempting to file this under "nice to have." Verify agents, good idea, moving on. That undersells it.

As agents start to act, and not just talk, the stakes change. When an agent moves money, signs something, or makes a commitment for you, the useful question is no longer whether it sounds human. It is who is answerable for what it just did. An accountable agent has an answer. A mimic does not.

This does not mean every agent must be verified. Most will not be, and that is fine. Plenty of agents will keep running anonymously, and for a lot of uses that is perfectly reasonable. Verification is for the agents and services that take on responsibility, the ones where it matters that someone real stands behind the action. The point is not to fence everyone in. It is to make the difference legible when it counts.

We are not the sole verifier

Something worth saying plainly: enabling accountable agents does not mean we have to be the one that verifies everyone. We could not, and we would not want to. Confirming who billions of people and companies are, across many countries, is not a job for a single operator, and the world already has trusted ways for people and companies to prove who they are.

So the accountability behind an agent can be carried in different ways. Some of it we might help with. Some can come from partners, or from providers that already do this well. Some people can bring their own. What stays constant is the bar: however the accountability is established, it has to meet the same baseline before an agent counts as verified. People do not have to come to us for that, but they do have to clear it. This is about enabling a kind of agent to exist, not about becoming the world's verifier or the largest pile of personal data.

Where .agent fits

This is what we are building toward with .agent. Our intention is that a .agent name becomes a trust mark that carries this accountability by default, so that anyone operating a .agent domain can prove their verified status through the name itself. A .agent name would not only say that something is an agent. It would carry, built in, the fact that a real and accountable party stands behind it.

What we want to promise

Here is the commitment underneath all of it. We want to enable agents that are accountable for their actions, because the legal entity behind them is accountable, and because that accountability is discoverable.

In a world filling up with agents, that one property, being able to tell a representative from a mimic, is going to matter more than almost anything else we could build. It is worth getting right while it is still early.