What trademark law reaches onchain, and what it does not

A brand owner who discovers that a string matching the company’s mark is registered as an onchain top-level domain often arrives at the same instinct: we have the trademark, so we should be able to stop this. The instinct is not wrong, but it is incomplete. Trademark law reaches further than many assume, and less far than most hope. The gap between those two boundaries is where strategy lives.

This article separates three questions that are almost always blurred in the first conversation. Does a right exist? Does the conduct in front of you amount to infringement? Is there a remedy that actually works? Each question has a different answer onchain than it does in the traditional Domain Name System (DNS), the naming system that ICANN coordinates. Working through all three lets you set realistic expectations, brief your lawyers precisely, and choose a commercial path before committing to an enforcement route that leads nowhere.

What a trademark actually protects

A registered trademark gives its owner the exclusive right to use a sign in the course of trade for the goods or services covered by the registration. Two conditions drive most infringement analysis: the contested use must occur in the course of trade, and it must create a likelihood of confusion among the relevant public. Where a mark has a strong reputation, a use that takes unfair advantage of that reputation or damages it can also be caught, even without confusion. The threshold for that wider protection is high, and it still depends on commercial use of some kind.

Neither condition is automatic. A string of characters sitting in a wallet, with no product attached and no consumer ever looking at it, is not obviously “use in the course of trade” in any legal system. What brings conduct inside the protection a trademark offers is the connection between the sign and commerce: presenting it to customers, attaching it to goods or services, using it to route payments, or putting it on a website that offers something. Once that connection is present, the trademark analysis sharpens. Until it is, the analysis is more tentative than most brand owners expect.

Reputation-based claims go by different names. In the United States they are usually discussed as dilution; in the European Union they are discussed as the extended protection of marks with a reputation. The tests differ in their detail, but both ask how well known the mark is among the relevant public, and both still look for a sign that is being used in commercial life. A dormant TLD does not obviously meet that requirement. A TLD under which a business is being run is a different matter.

One more general point shapes everything that follows. Trademark rights are limited by territory and by the goods and services they cover. A mark registered for one class of products in one country is not a worldwide right to a word. Onchain names have no territory and no class. That mismatch is the source of most of the difficulties described below.

Holding a string is not the same as using it

Freename, to take one onchain registry, records a TLD onchain in the form of a token that sits in a wallet. That record is public, and a block explorer lets anyone look it up. Holding the token establishes who controls the string. By itself it does nothing that a consumer can see. No website is published under it by default. No payment is routed through it. No name has been issued under it.

This matters a great deal for the infringement question. In the traditional DNS, a trademark owner has the Uniform Domain-Name Dispute-Resolution Policy (UDRP), an administrative procedure that domain registrants accept through their registration contract. Under that policy, simply holding a domain name without using it can, in some circumstances, count toward a finding of bad faith. Onchain names sit outside the ICANN system, so the registration contracts that carry the UDRP are absent and the policy does not apply. The question therefore falls back onto general trademark law, and general trademark law is cautious about treating passive ownership of a string as use.

A helpful comparison inside trademark law is the registered mark that is never used. Many systems allow such a registration to be challenged for non-use after a period, on the reasoning that rights in a sign are tied to real commercial activity. The same reasoning points the other way here: a claim of infringement normally needs commercial activity by the other side. A wallet holding a TLD that has never issued a name, never resolved a site and never been presented to any customer is far from that threshold.

For a brand owner, this cuts both ways. A dormant .yourbrand in someone else’s wallet has probably caused little or no harm to customers, and that leaves room to approach the situation commercially instead of in a hurry. It also means the trademark, on its own, may give you little to work with while nothing is happening.

The picture can shift where the circumstances suggest that the string was chosen because of the brand. Knowledge of an earlier right, combined with a string that is identical or confusingly similar to a distinctive mark, can be relevant to arguments about bad faith or unfair competition in some legal systems, even without active use. How much weight that carries varies considerably from one country to another. It remains a full court claim, with the practical obstacles described below, and not the streamlined procedure that DNS practitioners are used to.

When active use under the TLD changes the picture

Use begins when the TLD starts doing things that customers encounter. Three kinds of use are worth separating.

Payment names. A name issued under an onchain TLD can stand in for a long wallet address, so that a payer sends funds to a readable name. A customer who pays pay.yourbrand is treating that name as a commercial identifier. If .yourbrand is not held by the trademark owner, and the payment name is presented to customers as the way to pay a business in the relevant sector, the use-in-commerce element starts to look satisfied. The confusion analysis then turns on the usual factors: how close the string is to the mark, how far the goods or services overlap, and how attentive the relevant customers are.

Websites. A name under an onchain TLD can point to a website that opens in browsers supporting that namespace. A commercial site operated under a name that matches a mark, offering overlapping goods or services, is the clearest case of contested use. In principle the trademark analysis is the same as for a site on a DNS domain. The enforcement analysis is very different, as the later sections show. One practical nuance applies: because an ordinary browser does not open onchain names without support, the audience that actually sees such a site may be narrow, and the size of that audience can matter to both confusion and damages.

AI agents and other identities. Names under a TLD can serve as identities for AI agents that deal with customers or counterparties. If an agent introduces itself to a customer under a name that matches a brand in the same sector, the questions of commercial use and confusion arise in the ordinary way. What is less settled is how established doctrine applies when the party being addressed is itself software. Counsel should be asked about this use specifically, because it does not map neatly onto the situations trademark law grew up with.

Across all three, the principle is one of degree. The more a name is used to deal with customers in a commercial setting, the closer the conduct comes to classic trademark infringement. The longer it sits dormant in a wallet, the less the trademark gives you.

There is also a question of who is doing the using. A TLD and a name under it can be in different hands. On Freename, issuing names under a TLD is reserved to whoever holds that TLD, but once a name such as pay.yourbrand has been issued, the person operating it may not be the person who holds .yourbrand. The party whose conduct amounts to use, and the party who merely controls the TLD above it, may attract different analyses. That distinction should be made before any letter is drafted.

Who the defendant is, and why that matters

Before enforcement can begin, there must be a defendant: a person or legal entity against whom a court can make an order. A wallet address is not a person. It is an identifier generated by cryptography, and nothing in it states who controls it or where they live. The person behind it may be anonymous, may act under a pseudonym, and may be anywhere in the world.

Connecting a wallet to a person usually takes one of two things. The first is voluntary disclosure, which is unlikely in an adversarial setting. The second is legal process directed at an intermediary, such as an exchange or a marketplace, that has verified the identity of its customers and has dealt with that wallet. In some countries a court can order an intermediary to disclose who is behind an account. In others the legal basis is less clear, the intermediary may sit outside the court’s reach, or no intermediary holds the information at all because the wallet has never touched one.

Some legal systems allow an action against a domain name itself where the person behind it cannot be found or brought before the court. Those mechanisms were built around the DNS, where a registry or a registrar with a known location keeps the record. Whether anything similar is available for an onchain TLD is a question for counsel in the country concerned, and it should not be assumed.

Identification is also a cost, and that cost arrives before the merits are even reached. A brand owner may spend substantial effort discovering who holds a token, only to learn that the holder is in a place where a judgment would be hard to enforce. This is why the defendant question deserves attention first, not last. It decides whether the remaining questions are practical or theoretical.

Which court, which law

Assume a defendant has been identified. The choice of forum raises its own difficulties. Trademark law is territorial: a registration in one country does not create rights in another, and a court in one country does not automatically have authority over conduct carried out through a network that has no single physical location.

A defendant generally cannot be brought before the courts of a country with which they have no connection. Which country’s law applies usually depends on where the contested use has its effect. A website that can be reached from many countries might, in principle, be challenged in several of them, but courts in many systems look for some sign that the activity was directed at their territory, such as the language used, the currency quoted or the customers served. A payment name used by one local community may point clearly to one country, with fewer alternatives if that country’s courts prove unhelpful.

Brand owners accustomed to the DNS often find this the hardest adjustment. The UDRP works across borders because it rests on contract and not on any one country’s courts. Without it, the cross-border reach has to be rebuilt case by case from national rules.

In practice you are choosing among imperfect options. One country may offer the most favorable trademark law, a second may have authority over the holder, and a third may be where an order could actually be enforced. Those three rarely coincide. The choice calls for counsel who understands both trademark litigation and the way the particular registry works, because the procedure differs from a DNS dispute at almost every step.

Remedies and their limits

Even a successful claimant faces a remedy problem with no real counterpart in the DNS. A court can order a person to do something. It cannot itself move a token from one wallet to another. A blockchain records transfers that are authorized by the key controlling a wallet, and it was not designed around the execution of court orders.

The table sets out where the three questions land in each system.

Three kinds of order are conceivable onchain, each with limits.

An order against the holder. If the holder is identified and within the court’s authority, a court can order them to transfer the token, to stop using it, or to pay damages. The result depends on the holder obeying. The court’s leverage is the sanction for disobedience, which is effective against a person with assets and a presence in that country, and much less so against someone who has neither. A holder who cannot be identified, or who lives where foreign judgments are hard to enforce, is in practice beyond this remedy.

An order against a platform. Marketplaces on which tokens change hands may have their own intellectual property policies, and a rights owner can use the complaint procedure a marketplace provides. A court might also order a marketplace within its authority to stop offering a token. Neither step changes who holds the token or removes it from the chain. What it affects is one venue for selling it. The practical effect depends on which marketplace is involved and where it is established.

An order against a registry. A trademark owner could ask a court to order the operator of the naming system to stop a name from resolving, or to cancel the token altogether. Two separate questions follow. The first is legal: whether a court in the relevant country has authority over that operator. The second is technical: whether the operator has kept any ability to alter a record once a name has been issued. Where the design leaves transfers in the hands of the holder alone, an order to transfer asks for something the operator cannot perform. Do not assume the answer for any registry. Read its terms and have counsel ask.

What cannot be compelled technically therefore sets the outer limit of every remedy. An order is only as useful as the ability of the person receiving it to carry it out.

Why platforms and registries matter more than courts in practice

Given those limits, what happens to a contested name in practice is often decided by the policies of intermediaries and not by a judgment. The pattern is familiar from the DNS, where registrars, registries and hosting providers handle a large share of trademark complaints without any court being involved.

Onchain, several kinds of intermediary can be relevant. The registry that issued the TLD has terms of use and may have a complaint procedure. Marketplaces have intellectual property policies and are usually companies established in an identifiable place. Wallets and browser extensions decide whether a name resolves for their users at all. Where a name points to a website, whoever serves the content may be reachable too, although content published on a distributed storage network such as IPFS is harder to take down than content on a single host.

Each of these stands in a different relationship to the contested use, carries different legal exposure and applies different rules. A complaint under a policy does not need an identified defendant, does not depend on choosing a forum, and can be made in ordinary business correspondence. Its limits are just as plain. A policy decision affects one venue or one resolver. It does not move the token, and the intermediary is free to interpret its own rules.

For that reason, approaching the relevant intermediaries in a coordinated way is usually worth considering before, or alongside, any court action. This is not a criticism of the technology or of any operator. It is a description of where practical control sits.

All of this presupposes that you know what exists. A brand that does not know which onchain names match its marks cannot assess whether any of them are in use, and use is the fact on which both the legal analysis and any policy complaint depend. Namespace Watch, the service that monitors the names of a brand onchain, addresses that need.

How this legal reality shapes the commercial options

The enforcement picture leads to three practical orientations. They are not mutually exclusive, and many brand owners combine them.

Negotiation

Where a TLD matching your mark is held by someone else, an agreed purchase can be more predictable than litigation. A claim requires a defendant, a forum and a remedy that can be carried out. A purchase requires a willing seller and a price. A holder who has no project for the name may well be open to selling it. Treating the matter as a transaction is not a concession on the legal merits. It reflects the fact that, for a dormant name, the merits may be thin and the cost of testing them high. Acquiring the onchain TLD that matches a brand is what the brand namespace recovery service is for.

Negotiation has its own limits. A price that seems unreasonable is not evidence of infringement, and a refusal to sell leaves you where you started. Your trademark position still matters in the conversation, because it defines what the holder could lawfully do with the name in your sector.

Monitoring and patience

Where the name is dormant, the proportionate response may be to watch. A TLD that is not being used is not confusing your customers. If use begins, the infringement analysis and the available complaints both become stronger, and they are stronger still if you can show exactly what was presented to customers and when. Monitoring serves that purpose: it tells you when the facts have changed, so that you act on evidence of use and not on the bare fact of registration.

Patience is not passivity. It works only if someone is assigned to review what monitoring reports, and if you have agreed in advance with counsel what kind of use would prompt a response.

Holding the TLD yourself

The most direct way to control what happens under .yourbrand on a given registry is to hold it. Because only the holder of a Freename TLD can issue names under it, a brand owner who holds the TLD decides which names exist there. A Freename TLD also does not come with a yearly renewal charge, so there is no recurring registry fee for keeping it.

State the limit honestly. Holding a TLD on one registry settles the question for that registry. It does not affect other naming systems, and it does nothing about similar strings. The page on trademark TLDs explains what such a TLD is and carries the free Trademark Name Check. The page for brands and legal teams gathers the material written for general counsel and brand protection teams.

Where the desk fits

Hush Partners is a private desk for onchain TLDs. An organization can mandate it to find and approach a holder, and the desk never names the seller of a given TLD. A standard mutual NDA applies to every exchange, one named partner follows the file from the first note to closing, and closing runs through onchain escrow on an aftermarket. None of this replaces legal advice. Lawyers and advisers who want to refer a file will find the page for advisers addressed to them.

Questions to put to your own counsel

The sections above identify the questions. The answers depend on your mark, your markets, the registry concerned and the countries involved. Work through the following with counsel before you choose a path.

1. What exactly does our registration cover?Establish which goods, services and territories the registration covers, and whether the mark would qualify for reputation-based protection in the markets that matter. Rights that look broad in one country may be narrow in another.
2. Is there use in the course of trade?Determine whether the TLD, or any name issued under it, is resolving to a website, receiving payments or appearing in front of customers. Ask how the answer changes the analysis in each country you care about.
3. Who is using it, and can they be identified?Separate the holder of the TLD from the operator of any name under it. Ask what routes exist to identify either one, and what each route is likely to demand in time and effort.
4. Which forums are realistic?Consider where the use has its commercial effect, where an identified holder is located, and where an order could be enforced. Ask whether those three point to the same place.
5. What can the intermediaries do?Ask which registry, marketplaces and resolvers are involved, what their policies allow, and what each is technically able to do. Compare that with what a court could order.
6. How does a purchase compare?Set the cost and uncertainty of proceedings against an agreed acquisition, and ask what your trademark position contributes to a negotiation.
7. What would make us act later?If you decide to wait, agree what kind of use would prompt a review, who watches for it, and what evidence should be preserved when it appears.

Questions people ask

Does filing a trademark application after an onchain TLD is registered help?

It can help at the margins. A registration is evidence of a right, and that right is relevant to a negotiation, to a complaint under a marketplace policy, and to any later claim. In most systems, however, a new application does not reach back before its own priority date, and the rules on earlier unregistered rights vary by country. It also gives you no mechanism to compel the transfer of an onchain TLD in the way a UDRP complaint can for a DNS domain. Whether and where to file is a decision for your trademark counsel.

Can a court order a smart contract to transfer or cancel a token?

A court addresses people and organizations, not software. A smart contract is a program stored on a blockchain that runs as written. A court can order the holder of a token to transfer it, and it may be able to order a registry operator to act if that operator falls within its authority and its contracts give it the technical means. Whether those means exist differs from one registry to another. The glossary defines the technical terms used here.

Is there a dispute procedure equivalent to the UDRP for onchain TLDs?

No single one. The UDRP is a contractual mechanism: it binds registrants because registrars accredited by ICANN include it in their registration agreements. Onchain registries do not operate inside that framework. A registry may have its own complaint or objection procedure, but such procedures are not standardized, and what they can decide varies. Check the terms of the specific registry before assuming that any comparable route exists.

What if the holder uses the TLD in a completely different sector from the trademark?

Infringement based on confusion depends on how similar the goods and services are, not only on how similar the signs are. A holder using .yourbrand for a business in a clearly unrelated sector presents a weaker case on that ground. Reputation-based protection may still be relevant if the mark is well known enough. Monitoring remains worthwhile, because a use can move from one sector to another.