Protect a brand in a new .ai gTLD launch: what panels actually decide

Protect a brand in a new .ai gTLD launch: what panels actually decide. UDRP and ccTLD domain recovery and defense across .ai. Email the firm to assess your cas…

A technology company spends years building a product name, registers the trademark, and then watches an AI-sector entrant snap up the matching .ai domain the moment a new delegation opens registration. The question is not whether that behavior is opportunistic. The question is which legal tool reaches it — and what evidence actually persuades a panel to act.

To protect a brand in a new .ai gTLD launch, a rights holder must understand that .ai operates as the ccTLD for Anguilla and does not itself administer a UDRP-equivalent procedure as of this writing; disputes over .ai domains therefore generally proceed through the UDRP before WIPO or another accredited forum where the registrar agreement incorporates that policy, or through applicable court action, rather than through the Uniform Rapid Suspension system, which applies only to new gTLDs delegated in ICANN's expansion rounds. The legal standard in either path centers on all three elements of Paragraph 4(a) of the UDRP, and the outcome turns on the quality of the bad-faith evidence the rights holder can produce.

This analysis covers the governing procedure for .ai, the URS as a comparative baseline for new-gTLD brand protection, the evidentiary patterns that decide these disputes, and the strategic choices available to rights holders approaching a .ai zone launch.

Is .ai a new gTLD, a ccTLD, or something else — and why does it matter?

The .ai zone is the official country-code top-level domain for Anguilla and was not delegated through ICANN's 2012 or subsequent new-gTLD rounds; it is a legacy ccTLD, not a new gTLD in the technical ICANN sense. That distinction carries immediate procedural consequences. The Uniform Rapid Suspension system — a lower-cost, faster mechanism introduced specifically for new-gTLD domains — does not apply to .ai registrations. Brands that encounter infringement in .ai cannot file a URS complaint.

What does apply? The answer depends on the registrar. Many registrars accredited to sell .ai domains require registrants to submit to UDRP-style dispute resolution as a term of registration. Where that condition is present in the registrar agreement, a complainant may file a UDRP complaint at WIPO or the Forum in the ordinary way. Where it is absent, or contested, the route shifts toward national court action — meaning litigation in the jurisdiction governing the registry or the registrant's domicile, handled with local litigation counsel in the relevant jurisdiction.

Rights holders approaching a .ai launch should therefore verify, before any domain is registered by a third party, whether their target registrar's current terms incorporate UDRP submission. That single fact governs the entire dispute pathway.

Why does any of this matter for brand protection strategy? Because a rights holder who assumes .ai works like .com — and that a UDRP complaint is automatically available — may lose weeks discovering otherwise. In our practice, we regularly advise clients in the technology and AI sectors to run a registrar-agreement audit before a product launch, not after a domain squatting event.

For a preliminary assessment of whether UDRP or another route applies to a specific .ai registrar, contact info@cognomenlaw.com.

How does the URS differ from the UDRP, and when does each tool apply to protect a brand in a new .ai gTLD launch?

The URS and the UDRP share surface features but differ in standard, remedy, and scope in ways that are decisive for any brand-protection plan. Understanding the contrast helps rights holders working across both new-gTLD and ccTLD zones, including .ai, choose the right instrument — or know when neither applies alone.

The URS was built for speed in new gTLDs. It applies exclusively to domains registered under gTLDs delegated through ICANN's expansion program. The remedy is suspension — the domain stops resolving — for the remainder of the registration term. There is no transfer. That ceiling is intentional: the URS imposes a higher evidentiary threshold, requiring the complainant to show infringement by "clear and convincing" evidence, precisely because the procedure is faster and the registrant's opportunity to respond is compressed. A brand owner who wins a URS obtains a suspended domain; the registrant retains nominal ownership and may apply to restore it.

The UDRP, by contrast, offers transfer or cancellation as its remedies, which is a materially better outcome for a rights holder who wants to operate the domain or prevent its future misuse. The standard — satisfying all three Paragraph 4(a) elements — is met on the balance of the evidence, not by a clear-and-convincing showing. That lower standard, combined with the transfer remedy, makes the UDRP the preferred tool in nearly every case where it is available.

For .ai specifically: because .ai is not a new gTLD, URS is unavailable. If the registrar agreement incorporates UDRP, a complaint at WIPO (filing fee starting at USD 1,500 for a single-member panel covering up to five domains) is the primary administrative route. If UDRP is not incorporated, court action is the only path. That path can reach monetary damages — something neither the UDRP nor the URS permits — but it takes substantially longer and costs considerably more.

The practical implication: a brand operating across a new-gTLD portfolio (say, .app or .tech) and the .ai zone simultaneously may be running a URS proceeding on one domain and a UDRP or court case on another, governed by different standards and yielding different remedies. We have advised registrants and complainants in exactly that configuration, and the evidence-assembly challenge is real.

What are the three UDRP elements and how do panels assess them in a launch context?

Paragraph 4(a) of the UDRP sets out three elements, all of which the complainant must prove. Each presents a distinct challenge in a new-launch scenario, where the chronology of the domain registration relative to the trademark matters enormously.

First element: confusing similarity. The panel compares the second-level domain to the complainant's mark. In practice, panels treat this element as largely mechanical: where the domain reproduces the mark exactly or with minor additions (a generic word, a number, a hyphen), confusing similarity is almost always found. In .ai-zone disputes, panels have noted that the zone itself carries connotations of artificial intelligence, which may amplify confusion where the mark is associated with technology products. This does not guarantee a finding — it is a factual assessment — but rights holders in the AI sector should document the way consumers associate their brand with that technology.

Second element: no rights or legitimate interests. The complainant must make a prima facie showing; the burden then shifts to the respondent to rebut. Panels look to the Paragraph 4(c) safe harbors: was there a bona fide offering of goods or services under the name before the dispute arose? Is the respondent commonly known by the domain? Is there a legitimate noncommercial or fair use? In a launch-day or sunrise-period registration by an apparent squatter, there is often no evidence of any prior use under the name. That makes the second element the strategic pivot point: a complainant who can show no plausible commercial reason for the respondent to hold the name — combined with an absence of prior use — typically prevails here.

Third element: bad faith in registration and use. This is where launch-context cases become complicated. The UDRP requires that the domain was registered and used in bad faith — a cumulative test. Panels apply the Paragraph 4(b) non-exhaustive factors: did the respondent register to sell the domain to the mark owner at a profit? To disrupt the complainant's business? To attract users by creating confusion? In a launch-day registration by an entity with no plausible connection to the name, the registration timing alone is a powerful indicator. The question is how much weight a panel gives to circumstantial evidence of intent when the domain resolves to a parking page or is simply held passively.

The passive-holding doctrine is the most actively contested area in this factual pattern. The consensus view among panels is that passive holding of a domain that corresponds to a well-known mark can constitute bad-faith use, even absent active deception. The contrary view — held by a minority of decisions — is that a complainant must produce concrete evidence of use that harms the mark or misleads consumers, and that passive holding alone is insufficient absent fame or distinctive character. Rights holders pursuing .ai disputes need to be prepared for that minority argument and to build a record that defeats it, not simply assume the consensus will hold.

What evidence actually decides the outcome in a .ai brand-protection dispute?

Evidence, not legal theory, determines panel decisions. In our practice, we have seen well-constructed complaints fail on thin evidence and technically imperfect complaints succeed because the factual record was overwhelming. The following categories of evidence are the ones that move panels.

Trademark registration and use predating the domain. A registration certificate alone is the floor, not the ceiling. Panels want to see the mark in active commercial use before the domain was registered. Product pages, advertising materials, press coverage, and sales data all help establish that the name had acquired meaning in the market before the registrant acted. The stronger the pre-registration use, the harder it is for a respondent to argue they were unaware of the mark.

Timing of the registration relative to the mark and any launch announcement. A domain registered within days of a product launch announcement, a trademark application publication, or a press release is among the strongest circumstantial indicators of opportunistic registration. Document the launch timeline meticulously. Preserving contemporaneous web captures and social-media posts about the product rollout gives the panel a chronology it can follow.

The respondent's conduct after registration. Panels scrutinize what the registrant does with the domain. A parking page displaying pay-per-click advertising in the complainant's product category is direct evidence of attempted commercial gain by confusion — one of the Paragraph 4(b) factors. A demand for payment, particularly a "five-figure buy-back demand," is equally damning. No conduct at all — the pure passive hold — triggers the passive-holding analysis described above and requires the complainant to build the inference from the mark's fame and the absence of any plausible legitimate use.

The respondent's pattern of registrations. Paragraph 4(b) of the UDRP identifies a pattern of conduct preventing mark owners from reflecting their marks in corresponding domains as a non-exhaustive bad-faith indicator. A registrant who has registered multiple brand-matching domains across zones — including in .ai alongside .com, .io, or .tech — presents that pattern evidence. Compiling WHOIS/RDDS history and cross-checking registrant details across zones is a routine part of case preparation in our practice.

In a recent matter involving a .ai zone domain and a US-registered AI-product trademark (autumn 2025), we assembled a timeline showing that the disputed domain was registered within 72 hours of the complainant's press-announced product launch, combined with evidence that the registrant held approximately a dozen other brand-matching domains across new gTLDs and legacy ccTLDs. The panel found bad faith established on the cumulative record without relying solely on the passive-holding doctrine.

What is the consensus view on passive holding, and where does the minority diverge?

Passive holding — where a domain resolves nowhere or parks with a generic error page, without any active commercial exploitation — has generated more UDRP jurisprudence than almost any other issue. The reason is structural: bad-faith use is part of the third UDRP element, and a domain that simply sits idle tests what "use" means.

The consensus view, reflected in a substantial body of panel decisions, is that passive holding can constitute bad-faith use where the mark is well known, where there is no conceivable legitimate use of the domain by the respondent, and where the respondent has provided no explanation for the registration. This approach — sometimes called the Telstra line in panel discourse — holds that the cumulative circumstances surrounding a passive hold can satisfy the use requirement as effectively as active misuse.

The minority position requires something more. A minority of panels has insisted that bad-faith use demands actual conduct — a website, a demand letter, a commercial transaction — and that inaction alone, even by a registrant who clearly has no use for the name, is not enough to satisfy the element as written. Under that reading, a purely passive hold of a .ai domain matching a complainant's mark would require the complainant to wait for active exploitation before filing.

Which view a panel applies can turn on the composition of a three-member panel and the particular facts. Rights holders who want to reduce exposure to the minority view should take proactive steps: monitor the domain for any change in DNS, preserve evidence of the registrant's other holdings, and consider requesting a three-member panel where the case is high value and the passive nature of the hold is likely to be contested. The cost differential — USD 4,000 for a three-member panel at WIPO versus USD 1,500 for a single-member panel — is often justified where the record is legally complex or where precedential clarity matters for a portfolio.

In our practice, we advise rights holders facing passive-holding fact patterns to build the record as if the minority view applies. If the consensus view holds, the extra evidence costs nothing. If the minority view is applied, that evidence may be the difference between transfer and dismissal.

If a .ai domain matching your brand is passively held by a third party, email info@cognomenlaw.com to assess whether the cumulative record supports a filing now.

How does a Sunrise period or Trademark Claims Notice service factor into a .ai launch strategy?

The Trademark Claims Notice service and Sunrise periods are mechanisms designed specifically for new-gTLD launches — and their applicability to .ai is limited but worth understanding as a comparative baseline for any brand team managing a mixed-zone portfolio.

In a new-gTLD delegation, the Trademark Clearinghouse (TMCH) acts as the central repository for validated trademark data. Rights holders who record their marks in the TMCH before a new gTLD's Sunrise period opens may register the matching domain in that gTLD before general availability. During the Claims period that follows, any registrant attempting to register a domain matching a TMCH-recorded mark receives a Claims Notice warning them of the trademark and requiring affirmative acknowledgment before the registration can complete.

.ai, as a legacy ccTLD, does not operate a TMCH-integrated launch process. There is no TMCH Sunrise or Claims Notice mechanism administered by the .ai registry. Rights holders cannot use TMCH records to preempt .ai registrations the way they can in new-gTLD launches. The monitoring and dispute tools available in .ai are, by comparison, reactive: watch services, WHOIS/RDDS alerts, and UDRP or court action after the fact.

That gap is significant. For a new-gTLD launch — .app, .ai-adjacent zones such as .bot or .tech, or any zone delegated through ICANN's expansion rounds — proactive TMCH recording and Sunrise participation are the most cost-effective brand-protection instruments available. They cost a fraction of a UDRP complaint and block infringement before it occurs. For .ai specifically, the equivalent strategy is a monitoring program combined with immediate UDRP filing preparation when an infringing registration is detected.

Rights holders managing portfolios that include both new gTLDs and .ai should maintain both strategies in parallel. Our work with technology-sector brand owners regularly involves a TMCH Sunrise strategy for the new-gTLD components of a launch and a real-time monitoring protocol for the .ai zone.

What is the realistic decision matrix: UDRP, URS, or court action for a .ai domain?

The right route depends on the zone, the registrar's terms, and what the rights holder actually needs as an outcome.

If the domain is a .ai registered through a registrar whose agreement incorporates UDRP, and the rights holder wants the domain transferred, a WIPO UDRP complaint is generally the fastest and most cost-effective administrative path. The WIPO filing fee starts at USD 1,500 for up to five domains on a single-member panel, and a standard case resolves in approximately two months. Legal fees for a straightforward single-domain complaint are additional and typically fall in the market range of USD 3,000 to USD 7,000 for routine matters.

If the registrar agreement does not incorporate UDRP — which remains possible for some .ai registrars — court action is the only path to a forced transfer. This route takes substantially longer, costs more, and requires local litigation counsel in the relevant jurisdiction (Anguilla or the registrant's domicile, depending on the available theory of jurisdiction). It has one compensating advantage: courts can award monetary damages, something administrative panels cannot.

If the infringed domain is a new gTLD (not .ai, but a zone like .app or .io), URS is available as a faster, lower-cost suspension mechanism. URS does not transfer the domain. Where the brand needs to operate the name — not merely suppress it — UDRP remains the superior tool even in new-gTLD zones, because the transfer remedy is available and the evidentiary standard, while still demanding, does not require the "clear and convincing" showing that URS imposes.

If the rights holder faces registrations across both .ai and one or more new gTLDs simultaneously — a pattern we have seen with technology product launches in early 2025 — the strategic answer is to file UDRP for the .ai domain (where UDRP is available) and URS for the new-gTLD domains if speed of suppression matters more than transfer. Where transfer is the goal across all zones, UDRP for each is the path, with the zones potentially consolidated into a single complaint if the registrant of record is the same across all of them.

A complaint covering multiple domains requires that the registrant of record be identical. Confirm that before filing; a multi-domain complaint that fails the common-registrant test can be dismissed as procedurally defective.

What do panels decide when the complainant's mark postdates the domain registration?

A rights holder whose trademark was filed or registered after the disputed domain was registered faces a structural obstacle under the UDRP. The standard bad-faith analysis assumes the registrant knew of the mark at the time of registration — which is difficult to assert when the mark did not yet exist.

Panels have addressed this in two ways. The first approach holds that the complainant must show trademark rights that predate the domain registration, and that a post-registration mark cannot retroactively establish bad faith at the time of registration. This is the dominant view. Under it, a rights holder whose domain was registered before the trademark application was filed should expect a significant challenge on the third UDRP element and should consider whether court action — which can assess conduct over time, not only at the moment of registration — is a better-suited vehicle.

The second, minority approach looks to whether the complainant had protectable common-law or unregistered trademark rights in the name before the domain was registered — rights that arose from use, not from formal registration. Panels applying this view have found that extensive pre-application use of a product name in commerce can establish trademark rights for UDRP purposes even before a registration certificate issued. For AI-sector companies that built a brand identity and customer base before formalizing trademark protection, this doctrine matters. But the evidence burden is high: the rights holder must demonstrate that the name functioned as a mark — distinctively identifying a single commercial source — before the domain was taken.

In a late-2024 matter involving a .ai domain and an AI-software product, we advised a client to assemble unregistered-rights evidence — three years of product releases, press coverage, and documented customer use of the name — to support a common-law rights argument as a fallback to the formal registration date. The file is ongoing, and no outcome is certain; we raise it to illustrate that the evidence strategy differs when the chronology is unfavorable.

What does the respondent-side analysis look like, and when does RDNH become relevant?

Not every .ai UDRP complaint is filed by a legitimate rights holder against a squatter. The reverse scenario — a large brand using the UDRP as a strategic weapon against a legitimate registrant — is a real pattern, and panels have the authority to declare Reverse Domain Name Hijacking where the complaint was brought in bad faith to deprive the registrant of a legitimately held domain.

An RDNH finding carries no monetary penalty under the UDRP. It is a reputational sanction: the panel on record finding that the complainant filed an abusive complaint. But the reputational weight matters. RDNH findings are published in WIPO and Forum databases, indexed publicly, and available to future panels assessing a complainant's conduct. For a public company or a brand with an ongoing enforcement program, an RDNH finding is a material event that general counsel will want to avoid.

When does RDNH risk arise in a .ai launch context? Primarily in three situations. First, where the complainant's mark postdates the domain registration by a meaningful period and the complainant files anyway without addressing the chronology. Second, where the registrant holds the name for a demonstrably legitimate reason — a personal name, a generic word in common use, or a prior business connection to the term — and the complainant frames the complaint as if those facts did not exist. Third, where the complaint makes factual representations about the registrant's conduct that are not supported by evidence, and a panel finds that the complainant must have known they were unsupported.

We regularly act on the respondent side in UDRP proceedings, including cases involving .ai domains where the claimed rights did not in fact encompass the registrant's use. In our assessment, a respondent facing a .ai UDRP complaint should evaluate whether an RDNH argument is available within the first days of receiving the complaint, not as an afterthought at the end of the response.

Related at COGNOMEN

• URS and new-gTLD dispute services: suspension, defense, and forum selection across new zones
• TMCH Claims Notice and Sunrise services: protect your mark before a new gTLD launch opens
• Domain assignment agreements: structuring a transfer when the dispute resolves by negotiation