In brief. This is the what-to-do-about-it companion to our doctrinal analysis of metaverse trademark law. Where Trademark Challenges in the Metaverse and Virtual Goods explains what the law is after the MetaBirkins verdict and the Supreme Court's Jack Daniel's decision, this article explains how a brand actually defends itself: how to build a virtual-goods portfolio under the USPTO's overhauled 2025 fee structure; how to write specifications that survive examination; how likelihood-of-confusion and trade-dress doctrine bend in virtual environments; and—above all—how to enforce against counterfeiters, gray-market sellers, and pseudonymous minters who hide behind wallet addresses and offshore marketplaces. It is not legal advice. Specific situations require individualized analysis from qualified counsel.
A handbag that weighs nothing
A consumer in Ohio spends roughly $4,000 to put a Gucci handbag on her Roblox avatar. The bag weighs nothing. It cannot be carried, resold at a consignment shop, or handed to a daughter someday. It exists only as a rendering inside one company's servers, summoned by code and visible to anyone who loads the same digital room. And yet the buyer paid more than many people spend on a physical luxury bag, because the thing she actually bought was never the leather. It was the brand—the signal, the status, the association with a house that has spent a century teaching the world what its name means.
That transaction, multiplied across hundreds of millions of users, is why trademark law has been dragged into virtual worlds it was never built for. The legal questions are now well known, and we have answered the doctrinal ones elsewhere: Can a registration for physical handbags reach a virtual one? When is an NFT protected expression, and when is it infringement? Those are the is questions. This article tackles the do questions, which are harder and more urgent for any brand with money on the line. How do you build a portfolio that protects code-handbags? How do you find the counterfeiters when there is no factory to raid and no shipping container to seize? How do you enforce against a seller who is a string of hexadecimal characters operating a smart contract that runs itself?
To keep the strategy concrete, this article follows Northwind Athletic, a hypothetical streetwear and footwear brand. Northwind has strong physical-goods registrations built up over two decades, a signature sole pattern customers recognize on sight, official virtual drops on a major gaming platform, an emerging NFT program, and a chronic counterfeiting problem on secondary marketplaces. Northwind's predicament is the one facing every digital-first brand at once—register, monitor, and enforce in an environment where the infringers may be anonymous, offshore, and beyond the reach of any single takedown. Everything Northwind does should be read as a worked example, not legal advice. (Our companion piece follows a different hypothetical—Cordova Maison, a luxury house confronting the doctrine; the two articles are designed to be read together.)
Why the doctrine forces the strategy: MetaBirkins and Jack Daniel's in one breath
You cannot design an enforcement program without knowing what a court will reward, so begin with the two decisions that set the strategic backdrop. We treat the doctrine fully in the companion article; here is only what a brand manager needs to act on.
In Hermès International v. Rothschild, No. 22-cv-384 (S.D.N.Y.), Mason Rothschild minted 100 NFT images of fur-covered Birkin handbags, called them "MetaBirkins," and sold roughly $1 million worth, positioning the project as commentary on fashion and animal cruelty. Hermès sued for infringement under Lanham Act § 32 (15 U.S.C. § 1114), dilution under § 43(c) (15 U.S.C. § 1125(c)), and cybersquatting. In February 2023 a jury found for Hermès on all counts and awarded $133,000; Judge Rakoff entered a permanent injunction in June 2023. The verdict did not hold that NFT art is categorically unprotected. It turned on Rothschild's commercial framing (his own messages called the project "a gold mine"), survey evidence of an 18.7% net confusion rate, and—decisively—his use of "MetaBirkins" as a brand name rather than a descriptive title. The appeal, argued before a visibly divided Second Circuit panel in October 2024, remains undecided as of mid-2026.
Layered on top is Jack Daniel's Properties, Inc. v. VIP Products LLC, 599 U.S. 140 (2023), in which the Supreme Court held that the speech-protective Rogers v. Grimaldi test does not apply at all when an accused infringer uses a mark as a source identifier for its own goods. In that posture, ordinary likelihood-of-confusion analysis governs. The Ninth Circuit then applied the same logic to NFTs in Yuga Labs, Inc. v. Ripps, 144 F.4th 1137 (9th Cir. 2025), confirming that pointedly confusing NFT uses of another's marks fall outside Rogers protection.
For a brand like Northwind, four operational lessons fall out of these cases, and they shape everything below:
The NFT or "virtual" label confers no immunity—the medium is irrelevant to liability. The danger zone is source-identifying use: an infringer who deploys your mark as a brand for its own product is exposed, while a genuine commentator who depicts your product within a differently branded work may not be. Contemporaneous evidence wins—Hermès prevailed because it had the defendant's own words and a clean confusion survey, which means monitoring and documentation are not housekeeping but the core of enforcement. And enforcement must be calibrated: the same letter you send a counterfeiter could turn a genuine artist into a sympathetic First Amendment defendant and a public-relations liability.
Hold those four lessons. The rest of this article is, in effect, how to operationalize them.
Building the portfolio: classification, specifications, and the 2025 fee overhaul
Why a Class 18 registration may not save a virtual handbag
The Nice Classification organizes registrations into 45 classes (34 for goods, 11 for services) and has stretched before—Class 9 absorbed software, Class 42 absorbed software-as-a-service—but virtual goods pose a deeper categorization problem. A virtual handbag is not a Class 18 handbag. It is code that renders a handbag inside a platform; the purchaser cannot carry it, cannot use it outside that environment, and owns nothing tangible. Yet consumers treat it as a handbag and associate it with handbag brands. The prevailing view, adopted by most offices, is that virtual goods are distinct products requiring their own protection, so a brand that relies solely on its physical-goods registrations is exposed in the precise market where unauthorized uses are exploding.
The classification practice is now reasonably settled in outline. The Nice 12th Edition (in force January 1, 2023) first explicitly recognized NFT-related goods, adding "downloadable digital files authenticated by non-fungible tokens" to Class 9. The 13th Edition (2024) refined the logic: an NFT is classified by reference to the underlying asset it authenticates, and "virtual goods" are generally treated as digital content in Class 9, with the applicant required to specify the content. A full digital-first program typically reaches well beyond Class 9, however, into Class 35 (retail-store services featuring virtual goods, online marketplaces) and Class 41 (entertainment services, providing online non-downloadable virtual goods for use in virtual environments). For the underlying mechanics of how classes work and why getting them right matters, see our overview of USPTO trademark classes.
For Northwind, that means the portfolio is not one virtual-goods filing but a coordinated set: Class 9 for downloadable virtual sneakers and apparel; Class 35 for an online store and marketplace selling them; and Class 41 for the in-game, non-downloadable experiences and the entertainment services that surround the drops. Each addresses a different commercial activity, and a gap in any of them is a gap an infringer can exploit.
Specificity is now mandatory—and now expensive
The USPTO's Trademark Manual of Examining Procedure (TMEP) requires identifications definite enough to identify the goods. "Downloadable virtual goods," standing alone, is too indefinite and draws a refusal. The examiner wants the equivalent of: "downloadable virtual goods, namely, computer programs featuring footwear and clothing for use in online virtual worlds." The EUIPO imposes the same demand, treating generic "virtual goods" claims as insufficiently clear, and the USPTO's guidance specifically asks applicants to indicate where virtual goods are used ("in online virtual worlds") and whether they are downloadable.
Two things make this more than a drafting nicety. First, an overbroad specification invites refusals and delay. Second—and this is the change that reshaped filing strategy in 2025—vague or custom language now costs real money.
Effective January 18, 2025, the USPTO scrapped its long-standing two-tier electronic-filing system. Gone are TEAS Plus ($250 per class) and TEAS Standard ($350 per class). In their place sits a single base application fee of $350 per class, available only when the applicant selects pre-approved descriptions from the Trademark ID Manual, layered with new surcharges that bite hardest on exactly the kind of bespoke language virtual-goods filings often need:
- a $100-per-class surcharge for an application missing required information;
- a $200-per-class surcharge for using free-form (custom) goods/services descriptions instead of the ID Manual; and
- a further $200 for each additional 1,000 characters of free-form text per class.
Madrid Protocol filings designating the U.S. under Section 66(a) rose to $600 per class. The practical effect is that a brand entering bespoke virtual-goods language can find a single custom-identification class costing $550 or more, and a long, hand-drafted specification can run higher still. The new structure quietly recreates a two-tier outcome through surcharges—and it rewards precisely the discipline virtual-goods filings require: ID-Manual-conforming descriptions, filed complete at the outset. Where the ID Manual already contains a serviceable virtual-goods entry, use it; where it does not, weigh the surcharge against the value of bespoke breadth. A comprehensive multi-jurisdiction virtual-goods program can still run well into five or six figures in government fees alone before counsel time, so the savings from disciplined drafting compound across a portfolio.
For the full mechanics of filing—from clearance through specimen and Statement of Use—see how to file a trademark application with the USPTO and our federal trademark application checklists. And because rights begin with availability, a virtual-goods program should start with a real trademark clearance search across Classes 9, 35, and 41, not just the physical-goods classes the brand already owns.
Timing, use, and the intent-to-use trap
U.S. rights depend on use in commerce, and that principle does not soften because the goods are digital. A brand that wants to reserve a mark before a virtual launch files an intent-to-use application under Section 1(b). But ITU rights are not free-floating reservations: once the USPTO issues a Notice of Allowance, the applicant must file a Statement of Use within six months, extendable in six-month increments up to a maximum of 36 months. Miss the window and the application goes abandoned. A speculative virtual-goods filing made years ahead of an actual drop therefore carries a live abandonment risk that has to be managed against a development calendar. (We cover the mechanics and pitfalls in intent-to-use trademark applications, and the strategic timing question more broadly in when to trademark your brand.)
There is also a subtler use trap specific to virtual goods: what counts as a specimen? A screenshot of a virtual sneaker displayed for sale inside a game, with the mark visible and a purchase mechanism present, may suffice; a mere concept rendering on a pitch deck will not. Northwind should be capturing real, dated point-of-sale evidence from its drops—the same evidentiary discipline, as it happens, that wins enforcement actions later.
Geography: rights are still territorial in a borderless market
Trademark rights are national, even when commerce is not. The Madrid Protocol streamlines but does not eliminate country-by-country prosecution, and for digital goods the priority markets cluster predictably: the United States, the European Union, China, South Korea, and Japan. China deserves particular attention because its courts have begun applying trademark protection to virtual goods aggressively—in a 2024 decision, the Hangzhou Intermediate People's Court found that virtual vehicle skins in an online game infringed a mark registered for physical vehicles, reasoning that consumers could perceive a source connection across the physical-virtual divide. South Korea's office (KIPO), by contrast, presumes virtual goods and their real-world counterparts dissimilar. That divergence is not academic. It means Northwind cannot assume its physical registrations will be treated as related to virtual goods everywhere, so the only safe strategy is to file affirmatively for virtual goods in each priority market rather than relying on cross-category protection that exists in some offices and not others.
Likelihood of confusion when the "goods" are pixels
Infringement under Section 32 turns on likelihood of confusion, assessed through the multi-factor tests the circuits have used for decades—in the Second Circuit, the Polaroid factors from Polaroid Corp. v. Polarad Electronics Corp., 287 F.2d 492 (2d Cir. 1961); in the Ninth, the Sleekcraft factors. The factors are familiar—similarity of the marks, strength of the senior mark, proximity of the goods, evidence of actual confusion, the junior user's intent, the quality of the goods, and consumer sophistication—but several of them behave strangely when the goods are pixels. (For how courts weigh these factors in practice and at summary judgment, see navigating the maze of trademark confusion and the Polaroid factors on summary judgment.)
The proximity factor is the one most transformed, and it usually helps the brand owner. Traditional analysis treats handbags (Class 18) and software (Class 9) as commercially distant—different shelves, different buyers, different trade channels. But a virtual handbag exists for the sole purpose of replicating the experience of a physical handbag. The psychological and commercial distance between Northwind's real sneakers and a virtual sneaker for an avatar is far smaller than the two-classes-apart appearance suggests, because the virtual product trades on exactly the associations the physical product built. The classification gap and the confusion gap point in opposite directions.
Bridging the gap—whether the senior user would naturally expand into the junior user's market—has flipped in brand owners' favor almost overnight. A decade ago, a court might have doubted that a footwear company would move into selling code. Today, major brands are visibly expanding into virtual goods, and surveys indicate that most metaverse purchasers expect established brands to have a virtual presence. That expectation is itself evidence: if consumers assume the real brand is behind the virtual product, the gap is already bridged in their minds. Nike, Inc. v. MSCHF Product Studio, Inc., No. 22-cv-1606 (E.D.N.Y. 2022)—the "Satan Shoes" dispute—found custom sneakers within Nike's natural zone of expansion, though it involved physical goods; whether identical bridging analysis governs purely virtual products is still being worked out, and proving it will often require expert testimony on industry trends and consumer expectations.
Consumer sophistication cuts in genuinely contradictory directions, which is why it is dangerous to assume. On one hand, some virtual goods command luxury-level prices—the $4,000 Roblox bag from our opening is real-world reporting, not hypothetical—and high prices usually imply careful buyers. On the other, metaverse demographics skew younger and may transact impulsively inside an immersive environment engineered to reduce friction. Northwind's counsel should resist a tidy story here and instead develop the actual purchasing context for the specific goods at issue.
A worked hypothetical. Suppose a third party sells "virtual NW Air" sneakers for avatars on a marketplace, using Northwind's word mark and its signature sole pattern, and prices them just below Northwind's official drops. On proximity, the virtual sneakers are about as close to Northwind's real product as goods can get—they are digital copies of the same product. On bridging the gap, Northwind already runs official drops, so the expansion is not foreseeable but accomplished. On intent, the below-official pricing and the lifted sole pattern suggest free-riding. The one soft factor is sophistication, and even there a confusion survey of actual marketplace buyers would likely settle it. This is the easy case—source-identifying commercial use—and it is the case Jack Daniel's and Yuga Labs tell you to pursue without hesitation.
Trade dress in virtual environments
Trade dress—the overall look and feel of a product—offers both an opportunity and a puzzle in virtual contexts. Physical trade dress requires distinctiveness and non-functionality. Consider Louboutin's red soles, protected as trade dress since Christian Louboutin S.A. v. Yves Saint Laurent America Holding, Inc., 696 F.3d 206 (2d Cir. 2012). When a virtual shoe carries a red sole, the mark arguably serves the same source-identifying function it does in the physical world—users recognize Louboutin by the red, exactly as intended.
The functionality analysis, though, may run differently when goods exist only as images. In the physical world, courts deny trade-dress protection to features that are functional—features that affect cost or quality or that competitors need to compete. A virtual shoe has no utilitarian function at all; it cannot be worn, cannot wear out, and cannot perform. That could make virtual trade dress easier to protect, because the functionality bar that defeats many physical claims may simply not apply. It also raises a novel question: is a feature that is functional on a physical product (say, a tread pattern engineered for grip) still "functional" when copied onto a virtual sneaker that needs no grip? The USPTO has accepted trade-dress registrations for virtual goods, but examination practice is still maturing, and the answers are not settled.
The strategic point for Northwind is concrete. Its signature sole pattern and silhouette are valuable precisely because they are recognizable, and recognizable features are what infringers copy first in virtual environments, where lifting a visual is trivial. Northwind should evaluate whether its existing trade-dress protections reach virtual reproductions or whether separate virtual filings are advisable—and it should do that before the knockoffs appear, because trade-dress claims are far stronger when the brand can show a consistent, policed visual identity. The deeper doctrine of when product configuration earns protection is treated in our analyses of trade dress for product design and configuration and balancing brand identity and functional design.
The enforcement core: building a calibrated program
Registration is the easy part. Enforcement is where digital-first commerce actually tests a brand, and where the gap between brands that engage early and brands that wait becomes visible. A serious program has four moving parts: monitoring, triage, the response toolkit, and documentation. Northwind needs all four running before it ever sends a letter.
Monitoring: you cannot enforce what you cannot see
Traditional trademark watch services scan registries and a slice of the web. They do not, by default, surf Roblox marketplaces, crawl NFT collections on OpenSea, or index in-game item shops in Fortnite or Decentraland. A digital-first brand's monitoring has to reach the places the goods actually live. That means watching major NFT marketplaces and their secondary listings; monitoring gaming and metaverse platform stores; tracking new mints associated with the brand's name or visual identity; and following social channels where infringing drops are promoted. Specialized vendors now offer blockchain-aware brand monitoring, and the spend is justified by a simple logic borrowed from anti-counterfeiting practice: reacting to isolated reports—the street vendor the CEO happens to spot—is the worst way to police a brand, because it captures a random fraction of the problem and lets the rest entrench. Systematic market sweeps and continuous internet monitoring are the baseline, and they should feed a single intake channel so nothing falls through the cracks. Our broader guide to policing brands across the open web—cybersquatting, keyword advertising, social media, and screen scraping—is brand protection online.
Triage: matching the response to the use
Not every unauthorized use deserves the same response, and treating them alike is how brands lose both cases and goodwill. A workable triage runs along a spectrum:
Large-scale, source-identifying commercial use—a business minting "Northwind" sneakers at scale, or a counterfeiter selling fake virtual drops—warrants the full toolkit: a formal cease-and-desist, platform complaints, and litigation if the conduct continues. This is the MetaBirkins fact pattern, and it is the case you are built to win.
Small-scale use by an individual creator—one person selling a handful of unauthorized items—may call for a platform takedown and informal outreach rather than a federal lawsuit, which would be disproportionate and expensive.
Derivative artwork with a genuine commentary purpose is the danger zone in the other direction. Here the Rogers factors may protect the creator, and an aggressive letter risks manufacturing a sympathetic defendant. Evaluate artistic relevance and whether the use explicitly misleads before acting, and consider selective enforcement.
Counterfeits and knockoffs warrant aggressive action across every available channel—and, as the next section explains, the counterfeiting statutes hand brand owners remedies that ordinary infringement does not.
Small fan tributes may be candidates for tolerance or licensing rather than confrontation. Heavy-handed enforcement against enthusiasts carries real reputational risk, especially among the younger, digital-native consumers who see virtual worlds as creative spaces. The smarter move is often a clear brand-guidelines program that channels that energy into authorized creation—a theme we return to under licensing below.
A note on a recurring trap: enforce consistently within each tier. Trademark rights can erode through tolerated infringement, and a brand that ignores ten infringers and then sues the eleventh invites equitable defenses. We cover those in understanding equitable defenses—laches, acquiescence, waiver, and estoppel.
The response toolkit, from gentlest to sharpest
The first instrument is usually the cease-and-desist letter, and its tone is a strategic choice, not a formality. A letter to a commercial counterfeiter can be hard; a letter to a teenage fan should probably never be sent at all. Calibrate the demand, the deadline, and the threat to the actual target, and remember that any letter you send may be screenshotted and posted. Our practical guides—drafting a trademark cease-and-desist letter and sending and responding to C&D letters—walk through the mechanics, and the recipient's perspective is covered in responding to a trademark cease-and-desist letter.
The second instrument is the platform complaint. Even on "decentralized" infrastructure, the consumer-facing interface is almost always centralized, and that is the chokepoint (more on this below). A timely, well-documented complaint to a marketplace's brand-protection or IP team can achieve a practical takedown of the listing even when the underlying chain record is immutable.
The third is the domain remedy for the cybersquatting that frequently accompanies infringing drops. A brand can proceed under ICANN's Uniform Domain-Name Dispute-Resolution Policy (UDRP)—faster and cheaper, resulting in transfer or cancellation—or under the federal Anticybersquatting Consumer Protection Act, 15 U.S.C. § 1125(d), which offers damages but requires litigation. We cover the choice in how to file a UDRP complaint for domain name disputes.
The fourth is litigation, reserved for the cases that justify it—and here the brand should think hard about forum. U.S. federal courts offer broad discovery and strong remedies (including the counterfeiting tools below), which is one reason brands often prefer to sue where a U.S. nexus exists. The threshold work of deciding whether a case is worth bringing is its own discipline; see evaluating and assessing a civil lawsuit.
Documentation: the unglamorous thing that wins
Return to the first lesson of MetaBirkins: Hermès won because it had the defendant's own words and a clean confusion survey. Documentation is not back-office tidiness; it is the case. Northwind should maintain a centralized, searchable database of enforcement matters—date, the infringing product, the action taken, how the seller was identified, and any links between suspected wrongdoers. It should preserve dated evidence of each infringing listing (screenshots authenticated to survive an evidentiary challenge), of the brand's own first use and continuous use, and of any actual confusion it encounters. The chain-of-custody discipline that anti-counterfeiting programs use for physical seizures translates directly to digital evidence, and the brand that has it can move fast when a target finally surfaces. On the specific and frequently litigated question of getting web evidence admitted, see authenticating website screenshots as evidence in federal court.
The counterfeiting and gray-market arsenal
When the unauthorized use crosses from "infringement" into "counterfeiting," the Lanham Act stops being merely protective and becomes punishing—and digital-first brands should understand the difference, because the remedies are dramatically stronger and because virtual counterfeiting and physical counterfeiting increasingly travel together.
What makes a "counterfeit" special
The Lanham Act defines a "counterfeit" as a "spurious mark that is identical with, or substantially indistinguishable from, a registered mark" (15 U.S.C. § 1127). Note the word registered: the heavy artillery is available only to brands that actually registered their marks for the goods at issue—a direct reason to file for virtual goods rather than relying on common-law rights. The companion term, "counterfeit mark," used in the ex parte seizure provision (15 U.S.C. § 1116(d)), additionally requires that the mark be registered, in use, and registered for the goods or services associated with the counterfeit. That requirement is the whole ballgame for a virtual-goods counterfeiting claim: if Northwind never registered for "downloadable virtual footwear," its strongest counterfeiting remedies against virtual fakes may be unavailable, no matter how blatant the copying.
A counterfeit, importantly, need not be a slavish reproduction. Courts have found counterfeiting where defendants put non-genuine product in genuine containers, as in the refilled Coca-Cola bottles of U.S. v. Petrosian, 126 F.3d 1232 (9th Cir. 1997), and where reconditioned goods were sold under original labels without disclosure, as in Westinghouse Elec. Corp. v. General Circuit Breaker & Electric Supply, Inc., 106 F.3d 894 (9th Cir. 1997). The principle—that the law looks at the mark and the deception, not just the physical article—maps cleanly onto virtual goods, where a "genuine container" might be an official-looking storefront wrapping a fake.
The remedies that change the math
Three remedies make counterfeiting claims worth the registration investment.
Ex parte seizure. Under Section 34(d) of the Lanham Act (15 U.S.C. § 1116(d)), a brand owner can obtain a court order to seize counterfeit goods and the records relating to them without prior notice to the defendant—a powerful tool precisely because counterfeiters destroy evidence the moment they learn of a suit. The digital analogue is still developing, but orders freezing accounts, marketplace storefronts, and associated assets pending litigation are increasingly familiar in online counterfeiting cases.
Statutory damages. Proving a counterfeiter's actual profits is often impossible because counterfeiters do not keep honest books. Section 35(c) (15 U.S.C. § 1117(c)) lets a plaintiff elect statutory damages instead—between $1,000 and $200,000 per counterfeit mark per type of goods, rising to as much as $2 million per mark per type of goods where the counterfeiting was willful (the ceiling set by the 2008 PRO-IP Act). For a brand facing an anonymous, judgment-proof-looking infringer, the threat of a seven-figure statutory award changes settlement dynamics and supports default judgments.
Treble damages and fees. Where the plaintiff shows the defendant intentionally used a counterfeit mark, Section 35(b) (15 U.S.C. § 1117(b)) makes enhanced relief close to mandatory: treble the greater of profits or damages, plus reasonable attorneys' fees, plus (at the court's discretion) prejudgment interest. Courts have held that willful blindness—deliberately avoiding knowledge that a mark is counterfeit—suffices to trigger this enhancement, a doctrine with obvious application to marketplace sellers who look the other way. (The general standard for fee awards in ordinary, non-counterfeiting cases is narrower; see Lanham Act attorneys' fees under 15 U.S.C. § 1117(a).)
Customs recordation: stopping fakes at the border
For a brand like Northwind that still sells physical sneakers—and most digital-first brands have a physical side—recording federal registrations with U.S. Customs and Border Protection (CBP) is among the highest-leverage, lowest-cost moves available. Recorded marks let CBP detain and seize infringing imports, and CBP shares information about seized goods with the brand owner, feeding the enforcement database. CBP inspects only a sliver of incoming shipments and targets them by risk factors—prior seizure history, country of origin, type of goods, importer and shipper identity—so a brand that feeds CBP intelligence about suspect shipments materially improves interception odds. Anti-counterfeiting best practice is explicit: register key marks in every country of concern, record them with customs wherever permitted, audit those recordations periodically, and train customs and law-enforcement officers to tell genuine goods from fakes (coding, tags, or covert markings on packaging help). The virtual and physical channels are not separate problems; counterfeits frequently move through the same networks, and a coordinated program treats them as one.
Gray-market goods: genuine, but unauthorized
Counterfeits are fakes; gray-market goods (also called parallel imports) are genuine articles sold abroad under the real mark and then imported into the U.S. without the trademark owner's authorization, undercutting the authorized U.S. market. They are a distinct and stubborn problem because the goods are real, which neutralizes the usual "consumers are deceived about source" theory. Brand owners nonetheless have several statutory bases: infringement and unfair competition under Lanham Act §§ 32 and 43(a) (15 U.S.C. §§ 1114, 1125(a)); the importation bar of Section 42 (15 U.S.C. § 1124); and Section 526 of the Tariff Act (19 U.S.C. § 1526(a)). The most reliable path runs through material differences: where the gray-market product differs materially from the authorized U.S. version—different warranty, formulation, packaging, or quality—courts will find the kind of consumer confusion that supports a claim, an approach blessed in the line of authority following Lever Bros. Co. v. United States. The limits matter too: in Kirtsaeng v. John Wiley & Sons, Inc., 568 U.S. 519 (2013), the Supreme Court held that copyright's first-sale doctrine applies to copies lawfully made abroad, foreclosing the use of copyright to block gray-market imports of copyrightable goods—so the trademark theories carry the weight.
Why does gray market matter for a digital-first brand? Because the logic is migrating. Virtual goods are frequently region-specific—priced, localized, or feature-limited by market—and as cross-border resale of virtual items and accounts grows, the "genuine but unauthorized for this market, and materially different" pattern will recur in code. A brand that understands its gray-market toolkit in the physical world is better positioned to argue the analogues when virtual region-shifting arrives in force.
Enforcing against the decentralized and the pseudonymous
Here is the problem digital-first enforcement is famous for, and it is genuinely hard. Traditional online enforcement leans on identifiable, reachable intermediaries—platforms, registrars, payment processors—that comply with takedown demands to avoid secondary liability under Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844 (1982), which imposes contributory liability on a party that intentionally induces infringement or continues to supply a product to a known infringer. Decentralized technology breaks that model in four ways at once. The minting transaction is immutable on the blockchain and cannot be erased. The governing smart contract may run autonomously, with no operator to enjoin. The underlying images may sit on distributed storage with no central host. And the creator may be nothing but a wallet address, pseudonymous by design. Confronted with all four, a brand can find itself unable to identify the infringer, remove the content, stop secondary sales, or even determine which nation's law applies.
And yet—this is the part that gets lost in the "you can't enforce on the blockchain" hand-wringing—meaningful enforcement remains, because no system is perfectly decentralized, and the points that are not decentralized are exactly where pressure works.
Centralized interfaces are the primary chokepoint. The major NFT marketplaces—OpenSea, Rarible, Foundation—are ordinary companies with terms of service, content-moderation teams, and IP-complaint processes. They cannot rewrite the chain, but they can delist an infringing item from the interface where buyers actually see it, which removes the listing from commercial circulation for practical purposes. A brand with active monitoring and well-documented complaints can achieve real takedowns this way. The mechanics, and the wider questions of platform responsibility for user uploads, connect to the secondary-liability debates we cover in Section 230 reform and platform liability for user-generated IP infringement.
Payment chokepoints are the second. Pseudonymity ends where value is realized. To convert crypto proceeds into spendable money, an infringer typically passes through a regulated exchange subject to know-your-customer rules—reachable by subpoena or court order. The cash-out is the moment the mask slips.
Blockchain analysis is the third. Specialized firms trace transactions and link wallet clusters to real-world identities, and that work, combined with subpoenas to marketplaces and exchanges for account, IP-address, and payment records, frequently pierces pseudonymity when a brand devotes the resources. The blockchain's immutability cuts both ways: the very permanence that frustrates takedowns also means the evidence of infringement and of the money trail can never be deleted.
The honest takeaway for Northwind is that decentralized enforcement is feasible but resource-intensive, and that the dollars go furthest at the centralized interfaces and the cash-out points. Spreading effort thinly across every pseudonymous minter is a losing game; concentrating it on the chokepoints, with documentation ready, is how brands actually win.
Licensing, creator programs, and interoperability—turning a headache into a channel
Enforcement is necessary but expensive, and a brand that thinks only in terms of cease-and-desist letters will spend itself into exhaustion fighting a tide. The sophisticated move is to convert as much unauthorized activity as possible into authorized activity—to give the brand's most creative fans a sanctioned, brand-controlled channel, and to reserve hard enforcement for genuine bad actors.
Virtual-goods licensing requires adaptation well beyond traditional merchandise terms, and several provisions are genuinely new. Platform exclusivity: may the licensee create for all platforms, or only specified ones, and what happens as new platforms emerge? Secondary-sale royalties: many NFT systems pay creators a royalty on resales—though, as we examine in NFT marketplaces and secondary-sale royalties, those royalties are increasingly unenforceable at the protocol level and depend on contract and marketplace policy, so the license must allocate them explicitly rather than assume the chain will. Term and termination: NFTs persist on-chain after a license ends, so the agreement must address what happens to already-minted goods—a problem with no clean physical analogue. Content-moderation responsibility and quality control matter doubly, because trademark law requires a licensor to control the quality of goods sold under its mark or risk a "naked license" that abandons rights; the discipline is the same one we describe in our trademark licensing discussions, even when the licensed good is code.
Interoperability is the live frontier. The animating vision of the metaverse is goods that travel—the same Northwind sneakers worn in Fortnite, Decentraland, and Roblox—and although today's reality falls well short of that portability, initiatives like the Metaverse Standards Forum are building frameworks that could one day enable it. Licenses written now should anticipate it: specify whether platform-locked goods may migrate as interoperability matures, who controls cross-platform quality, and how royalties follow a good that moves. Interoperability is also an enforcement complication, because a takedown on one platform may not reach a portable good that has already migrated to another—one more reason to address it in the contract rather than after the fact.
The best-known authorized-channel model is Nike's .SWOOSH platform, which invites creators to design virtual products using official assets, with revenue sharing. It does two things at once: it reduces enforcement cost by drawing creative energy into a sanctioned space, and it builds community goodwill that pure suppression destroys. For Northwind, a creator program could convert its chronic counterfeiting headache into a controlled, branded channel—turning would-be infringers into licensees and turning enforcement from a cost center into a smaller, sharper instrument aimed only at the actors who refuse the legitimate path.
International coordination in a borderless market
A single virtual-goods transaction can involve a buyer in one country, a seller in another, a marketplace incorporated in a third, servers in a fourth, and chain records that exist everywhere and nowhere. That topology strains the territoriality principle at the heart of trademark law, and there is no treaty that resolves it cleanly. What exists instead are practical approaches that brands are converging on.
Marketplace-based enforcement pressures the incorporated, reachable operator to adopt and enforce brand-protection policies—useful precisely because the operator is findable even when its users are not. Targeted jurisdiction selection means suing where the defendant, the assets, or the marketplace can actually be reached, often favoring U.S. courts for their discovery powers and remedies. Coordinated multi-jurisdictional enforcement goes after genuinely global commercial infringers on several fronts at once, which is why the anti-counterfeiting playbook insists on centralized coordination of global activity—so that resources flow to where they are needed and intelligence accumulates rather than scattering.
Harmonization is coming slowly. WIPO working groups and EUIPO guidance are nudging offices toward common classification practice, and the October 2024 European Parliament Resolution on virtual worlds confirmed that EU trademark law "fully applies" to the metaverse. But significant divergence persists—most sharply the split between offices that presume physical-virtual relatedness (the USPTO and EUIPO lean this way) and those that presume dissimilarity (KIPO). Until that gap closes, a global brand has no choice but to build portfolios that protect across key markets under current frameworks while watching the harmonization efforts. The geographic logic of trademark rights, and why "borderless commerce" does not dissolve national borders for trademark purposes, is the subject of our deeper treatment in the geographic scope of common-law trademark rights.
What to watch next
Several developments will reshape this terrain, and a brand that watches them can stay ahead rather than react.
Generative AI can now produce virtual goods on demand—"a luxury sneaker in the style of [brand]"—which raises secondary-liability questions for the platforms running the models and design-defect-style theories against the developers. Courts have begun allowing claims against systems that can "recall and recreate" protected design elements. The questions overlap with the training-data disputes in copyright claims against generative AI and the data-acquisition questions in data scraping after hiQ v. LinkedIn.
Augmented reality will overlay virtual branded goods onto physical spaces, creating fresh confusion risks when users encounter unauthorized AR brand activations stitched into the real world.
Web3 identity systems could cut either way—easing enforcement by linking a repeat infringer's wallets to a persistent identity, or frustrating it by enabling more sophisticated anonymity.
And on the legal side, the pending MetaBirkins appeal, continued USPTO examination guidance, the fallout from the 2025 fee overhaul, and possible congressional attention to platform liability all bear close watching. The contested boundary between a person's virtual likeness and a brand's marks—avatars, digital doubles, deepfakes—is its own emerging field, treated in the right of publicity meets digital doubles.
Conclusion: act before the unauthorized uses entrench
Trademark law's integration with virtual goods is still early, and courts, legislatures, and industry will keep refining it. But the fundamentals do not waver: prevent consumer confusion, protect goodwill, ensure quality. Those imperatives apply with full force in virtual environments, where consumers care about authenticity and source exactly as they do with physical goods. The hard part is operational—adapting established tools to a category that is neither traditional merchandise nor pure software nor simple expression, but borrows from all three.
The brands that come out ahead will be the ones that engage proactively: building portfolios in Classes 9, 35, and 41 under the new fee regime, with ID-Manual-conforming specifications that keep costs down; monitoring across marketplaces and platforms that watch services miss; registering for virtual goods so the counterfeiting statutes' powerful remedies are actually available; recording marks with customs; running a calibrated enforcement playbook that distinguishes a swindler from a fan; and building licensing and creator frameworks that turn would-be infringers into authorized partners. The brands that wait for perfect legal clarity may find that the unauthorized uses have already entrenched, the goodwill already diluted, the secondary market already trained to treat fakes as ordinary.
MetaBirkins proved that trademark owners can prevail even amid uncertainty. The lesson for Northwind—and for every digital-first business—is that prevailing is a matter of preparation: a portfolio built before the launch, a record built before the dispute, and a strategy attentive to both the continuities and the genuine discontinuities between virtual and physical commerce. The handbag may weigh nothing. The brand still does.
Related Articles
Trademark Challenges in the Metaverse and Virtual Goods (doctrinal analysis) · NFT Marketplaces and Secondary-Sale Royalties · Brand Protection Online: A Strategic Guide for Businesses · USPTO Trademark Classes · How to File a UDRP Complaint for Domain Name Disputes · Drafting a Trademark Cease-and-Desist Letter · Navigating the Maze of Trademark Confusion · The Right of Publicity Meets Digital Doubles · Intent-to-Use Trademark Applications
This article is intended for informational purposes only and does not constitute legal advice. Metaverse trademark law, USPTO fees, and enforcement practice are evolving rapidly; readers should consult qualified intellectual property counsel regarding their specific circumstances. For assistance, please contact our intellectual property and technology practice.
Selected Authorities
Statutes and fees: Lanham Act, 15 U.S.C. §§ 1114, 1116(d), 1117(b)–(c), 1124, 1125(a), 1125(c), 1125(d), 1127; 15 U.S.C. § 1051(b)–(d) (intent-to-use, Statement of Use); Tariff Act § 526, 19 U.S.C. § 1526(a); Trademark Counterfeiting Act, 18 U.S.C. § 2320; Prioritizing Resources and Organization for Intellectual Property Act of 2008 (PRO-IP Act); USPTO Trademark Fee Final Rule (eff. Jan. 18, 2025) (eliminating TEAS Plus/Standard; $350-per-class base fee plus surcharges; Section 66(a) Madrid fee to $600/class).
Cases: Hermès International v. Rothschild, No. 22-cv-384 (S.D.N.Y. 2023), appeal argued Oct. 23, 2024 (2d Cir., pending); Jack Daniel's Properties, Inc. v. VIP Products LLC, 599 U.S. 140 (2023); Yuga Labs, Inc. v. Ripps, 144 F.4th 1137 (9th Cir. 2025); Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844 (1982); Polaroid Corp. v. Polarad Electronics Corp., 287 F.2d 492 (2d Cir. 1961); Christian Louboutin S.A. v. Yves Saint Laurent America Holding, Inc., 696 F.3d 206 (2d Cir. 2012); Nike, Inc. v. MSCHF Product Studio, Inc., No. 22-cv-1606 (E.D.N.Y. 2022); U.S. v. Petrosian, 126 F.3d 1232 (9th Cir. 1997); Westinghouse Elec. Corp. v. General Circuit Breaker & Electric Supply, Inc., 106 F.3d 894 (9th Cir. 1997); Kirtsaeng v. John Wiley & Sons, Inc., 568 U.S. 519 (2013); Lever Bros. Co. v. United States (gray-market material-differences line of authority); "G.PATTON" decision, Hangzhou Intermediate People's Court (2024).
Agency and international materials: USPTO Trademark Manual of Examining Procedure; U.S. Patent & Trademark Office and U.S. Copyright Office, Non-Fungible Tokens and Intellectual Property: Report to Congress (Mar. 2023); WIPO Nice Classification (12th ed. 2023; 13th ed. 2024); EUIPO and KIPO virtual-goods guidance; European Parliament Resolution on the Policy Implications of the Development of Virtual Worlds (Oct. 17, 2024); Practical Law, Protecting Against Counterfeit Trademarks and Gray Market Goods, Creating an Effective Trademark Anti-Counterfeiting Program Checklist, and Brand Protection Online; contemporary practitioner analyses (2024–2026) of the 2025 USPTO fee overhaul and metaverse enforcement. Filing fees, classification practice, and case postures change; confirm current details before relying on them.