Introduction
The metaverse is a shared virtual space. Users play, work, shop and attend events in it through avatars. Brands now sell virtual goods there, including digital sneakers, handbags and clothing for avatars. Many are sold as non-fungible tokens, or NFTs, which record ownership of a digital item on a blockchain. Users pay real money for them, often in digital currency. Trademark law was built for physical goods. It protects signs that identify the source of goods and services, and it depends on classification systems such as the Nice Classification, on use in trade and on geographic markets. Virtual goods strain each of these. They have no physical channel of trade, users in any country can reach them, and creators, platforms and users all take part in making and selling them. Courts have begun to examine how the doctrine applies. In Hermès International v. Rothschild, an artist created NFTs known as MetaBirkins, inspired by the Birkin handbag.1 The case turned on whether the artist’s expressive use of the mark was covered by the Rogers test,2 a test that the Supreme Court later narrowed in Jack Daniel’s Properties, Inc. v. VIP Products LLC.3 In Nike, Inc. v. StockX LLC, an NFT resale platform was sued over NFTs associated with physical sneakers,4 a case that raises the issue of platform liability. India has no precedent on trademarks in virtual goods, and the Trade Marks Act, 1999 does not mention them, although, as argued below, they may qualify as “goods” under it.5
A brand owner may hold a registration for physical handbags but have no apparent rights over virtual handbags. An artist may claim artistic freedom over an NFT that bears the name of a popular brand, and a platform may claim that it provides nothing more than a venue for transactions. Courts in the United States have resolved these issues unevenly, and Indian and EU law have said little about them. This paper asks whether the present trademark regime sufficiently safeguards brand owners and consumers, ensures freedom of expression and holds platforms accountable. The study is doctrinal and comparative: it examines the Trade Marks Act, 1999, Sections 32 and 43(a) of the Lanham Act, EU trade mark law and the TRIPS Agreement, and compares India, the United States and the European Union on classification, use in commerce, confusion, dilution, expressive use and intermediary liability. It is confined to trademark law, and the law is stated as at October 2026. Because India has little case law, the Indian discussion relies on statutes and analogies.
Literature review
1. Limongelli and Sposini (2025). The paper examines the differences between the American and European approaches to intellectual property problems in the metaverse, specifically trademark protection, copyright and NFTs.6 Its major contribution is to show that current legal doctrines encounter problems because digital assets circulate across borders. Its comparative approach is relevant to the present research, although its general examination of intellectual property leaves room for a dedicated examination of trademarks.
2. Jon and Park (2025). This comparative study in the Computer Law & Security Review focuses on the registration and protection of virtual goods and NFTs in the United States, the United Kingdom and South Korea.7 It supports the practical view that trademark owners need express registrations for their digital goods and services and cannot rely on a registration for physical goods. Its jurisdictions differ from those chosen for this paper (India, the United States and the EU).
3. Mangaraj (2026). The author applies existing trademark principles to NFTs and virtual goods through a discussion of infringement, dilution, cybersquatting, jurisdiction and enforcement.8 These issues are pertinent to the present study, especially to its analysis of remedies and reform, and the paper is most valuable as a source on enforcement.
4. Murray (2023). In “Trademarks, NFTs, and the Law of the Metaverse”, the author sets out the theories under US trademark law that may apply to creators of NFTs, such as infringement, dilution, false endorsement and false designation of origin.9 While the article provides a good doctrinal basis, its restriction to the United States leaves open the position in the other two jurisdictions.
5. Anwer and Hussain (2024). The paper discusses the difficulty of determining “use in commerce” in virtual economies and distinguishes the record on the blockchain from the rights in the asset.10 Its approach to borderless transactions is valuable for the discussion of territoriality, jurisdiction and use.
6. Draheim and Schulze, Hogan Lovells (2023). This note discusses the Court of Rome’s decision in Juventus, the German case law on trade marks and artistic freedom, and EUIPO’s approach to identifying virtual goods with the necessary clarity,11 an issue of great relevance to classification. Any view on the extent of protection for such goods, or for well-known marks, must nonetheless be examined within the relevant statutory provisions and the facts of the particular case.
7. Revilla and Bro, World Trademark Review (2023). The article reviews USPTO practice on filings for metaverse marks and the requirement to specify digital goods adequately.12 It illustrates the practical importance of classification and examination, although the decisions in particular cases cannot be generalised to all virtual goods.
8. Hyman and Blosser (2025), The Recorder. The authors discuss the appeal in Hermès v. Rothschild on the “explicitly misleading” element of the Rogers test and the impact of Jack Daniel’s on expressive use.13 The article is useful for tracking the development of US doctrine.
9. Gupta, Khurana & Khurana (2025). “Pixel Marks & Parallel Markets” takes the Indian perspective, discussing Indian trademark law after Hermès v. Rothschild and the poor fit between existing trademark categories and virtual goods.14 It points out the difficulty of fitting virtual goods into the classification scheme and raises questions about “use in the course of trade” in virtual markets. These points highlight the need for more precise guidance from the Indian Registry and for analysis of the legal position under the Trade Marks Act and the current Nice Classification.
10. Das, Jus Corpus (2026). The article states that India has no legal provision on trademarks in the metaverse and that disputes will have to be settled under the existing law.15 It identifies pseudonymous wallets as a problem in identifying the defendant and proving use of the mark. Its call for classification guidelines is relevant here, though it must be read together with the statute and case law.
Overall, the literature highlights several recurring concerns: classification, use in commerce, expressive use, cross-jurisdictional enforcement and platform liability. Yet most of these sources consider each issue individually or focus on one jurisdiction. Against this background, this paper considers India, the United States and the EU from a comparative perspective, analyses the impact of Jack Daniel’s on the Rogers test in disputes concerning NFTs, and suggests an approach to platform liability. Its novelty thus lies in the comparative discussion of these issues rather than in any claim that the literature has not addressed them.
Conceptual framework
A. Virtual goods and NFTs
Virtual goods are digital items used in online environments, such as avatar clothing, accessories, game skins and virtual land. A user can perceive, use or display them, but they have no physical embodiment. An NFT is a unique identifier stored on a blockchain that shows who owns a particular digital item. The token and the item should therefore not be confused: the token may only point to an image stored elsewhere. The paper considers three types of dispute.
1. NFTs associated with a physical item. In Nike v. StockX, the token referred to a physical sneaker stored in a vault.
2. NFTs that are digital items in their own right. In Hermès v. Rothschild, the token referred to an image of the handbag; in Yuga Labs v. Ripps, the token itself was the collectible.
3. Virtual goods used within a platform, whether or not a token is involved.
This categorisation matters because a token associated with a physical item resembles a receipt, while a standalone token resembles a product.
B. Trademark theory
A trademark indicates the origin of the goods or services it covers. It enables consumers to identify goods they can rely on and reduces their search costs. It also safeguards the owner’s goodwill, the value the owner has earned because its goods or services are of high quality.16 The dilution doctrine protects the distinctiveness of a famous mark even where consumers are not confused.17 Because marks are also part of language, trademark law places limits on its own protection. This paper examines four doctrines that arise from this theory: use in commerce, likelihood of confusion, dilution, and the free-speech defence for expressive use.
C. A test of adequacy
Whether existing doctrine is sufficient for virtual goods is assessed against five criteria.
1. The legal system protects consumers against source confusion.
2. The legal system protects the owner’s goodwill and the distinctiveness of its marks.
3. The legal system leaves space for expression.
4. The legal system assigns appropriate responsibility to platforms hosting sales.
5. The legal system has international reach.
D. How virtual goods differ from physical goods
Virtual products differ from real products in three respects, each of which challenges an underlying presumption of trademark law.
1. There is no physical channel of commerce. Trademark law presumes that products go into shops and onto shelves, whereas virtual products travel through wallets, marketplaces and avatars. Use in commerce and the classification of goods rest on this presumption.
2. There is no geographical restriction. Registration and enforcement are national, but a virtual product sold in one country is accessible in all others.
3. Several actors take part in production and sale. Creators, platforms and users are all involved, while trademark law presumes an identifiable seller, and pseudonymous wallets obscure the seller’s identity.
Current frameworks
A. International framework
Trademark rights are national; treaties set a common floor. The Paris Convention provides for national treatment and protects well-known marks under Article 6bis.18 TRIPS then sets minimum standards and incorporates the Paris rules.19 Four TRIPS provisions matter here.
1. Article 15(1) makes any sign capable of distinguishing goods or services eligible for registration. It does not exclude virtual goods.
2. Article 16(1) gives the owner the right to stop others from using identical or similar signs for identical or similar goods or services where confusion is likely, and confusion is presumed for identical signs on identical goods. Articles 16(2) and 16(3) extend protection to well-known marks, including for goods or services that are not similar where use would suggest a connection with the owner and harm its interests. This is the treaty basis for dilution-type claims over virtual goods.
3. Article 19 allows cancellation for non-use only after an uninterrupted period of at least three years, unless valid reasons for non-use are shown. A brand that registered for physical goods and never sold a virtual version may face this risk.
4. Article 41 requires enforcement procedures that permit effective action against infringement.
TRIPS has no provision on virtual goods, NFTs or platform liability.
The Nice Classification sorts goods and services into classes. Its 12th edition, which took effect on 1 January 2023, added “downloadable digital files authenticated by non-fungible tokens” to Class 9.20 The goods covered are the digital files, so a Class 9 registration may not reach the physical goods or other utility that the token also represents.21 EUIPO treats virtual goods as digital content in Class 9 and requires the applicant to name the type of digital item, so “NFT” alone is not accepted.22 Retail services for virtual goods fall in Class 35, while clothing sits in Class 25. A virtual jacket and a real jacket therefore sit in different classes.
WIPO administers the Paris Convention and the Nice system. Its 1999 Joint Recommendation on well-known marks is not binding.23 None of these instruments deals directly with virtual goods or platform liability.
B. India: the Trade Marks Act, 1999
Parliament passed the Act to bring Indian law into line with TRIPS. Because no section mentions virtual goods, the Act’s application to them rests on general words. Subject matter. Section 2(1)(zb) defines a trade mark as a mark capable of being represented graphically and of distinguishing the goods or services of one person from those of others, and a virtual good’s name or logo meets this test. The Act protects the use of a trademark on goods or services, and Section 2(1)(j) defines “goods” as anything that is the subject of trade or manufacture.24 Virtual goods are traded, so they may qualify, although commentators say that they sit poorly in the Nice classes the Registry uses.25
Infringement. Section 29(1) covers use in the course of trade of an identical or similar mark where confusion is likely. Section 29(4) protects reputed marks against use on dissimilar goods that takes unfair advantage of, or is detrimental to, the mark’s distinctive character or repute; this is the dilution provision.26 Well-known marks have added protection under Sections 2(1)(zg) and 11.27 The phrase “in the course of trade” raises doubt. A virtual sale made through an anonymous wallet, or offshore, may or may not meet it, and no Indian ruling answers the question.28
Section 30(1) protects use that identifies goods or services as those of the proprietor, if the use accords with honest practices and takes no unfair advantage of the mark, and Section 30(2)(a) protects descriptive use.29 The Act has no express exception for artistic expression or parody. The Delhi High Court has dealt with expressive use case by case, as in Tata Sons v. Greenpeace.30 Non-use. Section 47 allows a mark to be removed for non-use where five years have elapsed from its entry on the register without bona fide use up to a date three months before the application.31 India’s period is longer than the TRIPS minimum of three years, so a brand that registers for virtual goods without using them has a longer window than TRIPS requires but still faces removal. Section 134 lets a plaintiff sue where it resides or carries on business.32 That helps against cross-border virtual sales, but enforcement against foreign or anonymous sellers remains hard. The Act has no provision on virtual goods, NFTs or the metaverse, and commentators say that India has no landmark ruling on the point and that the Act must be stretched by interpretation.33
Case studies
A. Opening note on Lanham Act §§ 32 and 43(a)
The cases in this part arise under the Lanham Act. Section 32 protects registered marks: it bars use in commerce of a mark that is likely to cause confusion in connection with the sale or advertising of goods or services.34 Section 43(a) covers registered and unregistered marks and bars false designation of origin and false association. Section 43(c) protects famous marks against dilution by blurring or tarnishment, Section 43(d) bars cybersquatting, and Section 45 defines use in commerce. Courts test likelihood of confusion with multi-factor tests: the Second Circuit uses the Polaroid factors35 and the Ninth Circuit the Sleekcraft factors.36 Courts also read the First Amendment into the Act through the Rogers test.
B. Hermès v. Rothschild
Facts. Mason Rothschild sold 100 NFTs called MetaBirkins at about $450 each, each linked to an image of a fur-covered Birkin handbag.37 Hermès sued in January 2022 for infringement, dilution and cybersquatting, the last over the domain MetaBirkins.com.38
Pretrial rulings. At the pleading stage, Judge Rakoff held that the Rogers test applied, because the NFTs could be artistic expression, but that Hermès had plausibly alleged that the use was explicitly misleading.39 At summary judgment he again applied Rogers and found factual disputes on artistic relevance and on whether the use misled.40 He reasoned that buyers purchase an NFT to own the image attached to it, so the title MetaBirkins referred to the image as well as to the token. Hermès offered a survey showing 18.7% net confusion among likely NFT buyers.41
Verdict. On 8 February 2023 a jury found infringement, dilution and cybersquatting. It awarded $133,000, made up of $110,000 in profits and $23,000 in statutory damages.42 The jury also found that the First Amendment did not protect Rothschild. In June 2023 the court denied judgment as a matter of law and entered a permanent injunction,43 reasoning that the verdict showed that Rothschild had intentionally misled consumers.44
Appeal. Rothschild argued that Rogers required dismissal, and that Dastar bars trademark claims over intangible goods. Hermès argued that Rothschild used MetaBirkins as a brand and so fell outside Rogers.45 The Second Circuit heard argument on 23 October 2024.46 One report described Judge Pierre Leval as receptive to Rothschild’s reading of “explicitly misleading”, while the other judges on the panel echoed the district court’s concern that Rothschild was trading on Hermès’s goodwill.47 As at October 2026, when the law is stated, the Second Circuit had not decided the appeal.48
Analysis. Three points emerge. First, Hermès’s rights in a physical bag reached virtual handbags: the court treated the NFTs as covered by those rights, which, for the United States, answers part of the question whether protection for physical goods extends to virtual goods. Second, the court did not treat art and product as opposites; it found that the NFTs could be both. Third, the outcome turned on intent. The jury found that Rothschild meant to mislead, and a defence that depends on the defendant’s intent is hard to predict before trial.
C. Rogers v. Grimaldi after Jack Daniel’s
Rogers concerned a film title. The Second Circuit held that the Lanham Act does not reach the title of an artistic work unless the title has no artistic relevance to the work or explicitly misleads as to its source or content.49
In Jack Daniel’s v. VIP Products, VIP sold a dog toy called Bad Spaniels that copied the Jack Daniel’s bottle. The Ninth Circuit applied Rogers and ruled for VIP. The Supreme Court unanimously vacated that judgment. It held that Rogers does not apply when an alleged infringer uses a mark as a designation of source for its own goods; the claim then rises or falls on likelihood of confusion.50 The Court did not decide whether Rogers is sound in other settings. It also held that the noncommercial-use exclusion from dilution liability does not shield a parody used as a source identifier.
The Second Circuit applied this logic in Vans v. MSCHF, holding that Rogers did not apply to the Wavy Baby shoe and applying the Polaroid test.51
The NFT problem. A name like MetaBirkins does two jobs: it titles an image, and it brands a product line. Jack Daniel’s turns on which job dominates. Rothschild used MetaBirkins as the collection name, the domain name and the social media handle, and Hermès argued that this made it a source identifier. If that view prevails, Rogers drops out of NFT disputes in which a creator brands a collection with another’s mark, and the case then depends on confusion. In Yuga Labs, the Ninth Circuit rejected the First Amendment defence and sent the question of confusion back for trial.52
Assessment. In Hermès the artist obtained a Rogers instruction and still lost on intent. After Jack Daniel’s, an artist may lose the instruction as well, so that expression survives only inside the confusion analysis. That narrows the room for parody and commentary in virtual goods. The comparative analysis below asks whether the EU and India give more room, and the recommendations ask whether virtual goods need a specific safe harbour for commentary.
D. Nike v. StockX
Facts. StockX is a resale platform. It launched Vault NFTs, each of which tracked ownership of a physical sneaker that StockX held in storage and that the holder could redeem for the shoe.53 Nike sued in February 2022 in the Southern District of New York for infringement and dilution, alleging that the NFTs used Nike’s marks and images without permission. StockX answered that the tokens tracked ownership of physical goods and that no buyer would be confused as to source. In May 2022 Nike added claims that StockX had sold counterfeit shoes.
Rulings. On 4 March 2025 Judge Caproni granted Nike partial summary judgment on counterfeiting: StockX had sold 37 pairs of counterfeit Nikes, four to Nike’s investigators and 33 to a customer. The court refused Nike summary judgment on its other claims, including infringement, dilution and false advertising, and sent them to a jury. StockX had halted its Vault NFT program.54 The parties settled, and on 29 August 2025 the case was dismissed with prejudice; the terms were not disclosed.55
Analysis. The case gives less guidance on platform liability than its name suggests. First, the NFT claims never reached judgment. Second, the one ruling against StockX rested on StockX’s own sales of counterfeits. StockX was the seller, so its liability was direct and not secondary, and the ruling does not decide when a neutral marketplace is liable for what its users sell. Third, the Vault NFTs were receipts for physical goods and are a poor model for virtual-only goods.
The case does raise one useful question: if a token only tracks a physical shoe, may a brand owner still claim confusion over the token’s branding? No court decided it. For secondary liability, US law still rests on Tiffany v. eBay, which asked what a platform knew about specific infringing listings.56 The comparative analysis below builds the platform strand on that case, the Digital Services Act and Section 79 of the Information Technology Act, 2000.
E. Yuga Labs v. Ripps
Yuga created the Bored Ape Yacht Club NFTs. Ripps and Cahen made a near-identical collection using the same images and names, which they described as part of an expressive protest.57 The district court granted summary judgment to Yuga, and the awards reached about $9 million. On 23 July 2025 the Ninth Circuit held that NFTs are goods under the Lanham Act. It reversed the infringement and cybersquatting rulings because Yuga had not proved confusion as a matter of law, and remanded for trial; it rejected the nominative fair use and First Amendment defences. The parties settled in April 2026.58 For the Ninth Circuit, the case settles that a mark can protect an NFT as a good. It also shows that confusion must still be proved for virtual goods.
F. India: no decided metaverse case
India has no reported decision on trademarks in virtual goods or NFTs. Two Delhi High Court cases offer analogies. In Daimler Benz v. Hybo Hindustan, the court protected a famous mark against use on unrelated goods, a dilution-type result.59 In Tata Sons v. Greenpeace, the court dealt with the expressive use of a mark in an online game and refused an interim injunction.60 Neither concerns virtual goods.
Comparative analysis: US, EU and India
A. Overview of the US and EU frameworks
United States. The Lanham Act, outlined at the start of the case studies, protects registered and unregistered marks. Rights follow use in commerce. The Ninth Circuit holds that NFTs are goods. The First Amendment shapes expression defences through Rogers, as narrowed by Jack Daniel’s. Famous marks have a federal dilution claim. Platform liability is judge-made and rests on Tiffany v. eBay.
European Union. EU trade mark law has two layers. Regulation (EU) 2017/1001 creates the EU trade mark, which EUIPO registers and which applies across the Union.61 Directive (EU) 2015/2436 harmonises national trade mark laws.62 Rights follow registration, and a mark is open to revocation if it is not put to genuine use within five years.63
Article 9 lets the owner stop the use of identical or similar signs on identical or similar goods where confusion is likely, and Article 9(2)(c) gives the owner of a mark with a reputation a claim where use without due cause takes unfair advantage of, or is detrimental to, the mark.64 Article 14 limits the effect of a mark, for example for honest descriptive use and for use of the mark to identify or refer to the owner’s own goods.65 EUIPO accepts virtual goods in Class 9 if the specification is clear, and it rejects “NFT” alone. The first European NFT ruling came from the Court of Rome in 2022 in Juventus v. Blockeras.66 The court granted an injunction against NFT player cards that used the club’s marks. It held that the club’s Class 9 registration for downloadable electronic publications covered the cards, and it also treated the club’s marks as well known.67 Platform liability sits in the E-Commerce Directive, as read in L’Oréal v. eBay, and now in the Digital Services Act.68 The United States has the most NFT case law; the EU has a registration-based system with a reputation tier; and India has a statute and no case law. Together they span litigation-led and statute-led law, a range that suits the comparison this paper undertakes.
B. Comparison by issue
Classification. In the United States, the USPTO accepts virtual goods in Class 9 if they are specified. It registered METAJACKET for virtual jackets and objected to CYBERSNEAKER as descriptive.69 In the EU, Class 9 covers virtual goods and the applicant must name the digital item. India uses the Nice classes, but the Trade Marks Registry has issued no dedicated guidance on virtual goods,70 and commentators say the goods fit poorly. Section 2(1)(j) of the Act defines goods as anything that is the subject of trade or manufacture, which may cover them.71
Litigation has not required a separate virtual registration. In Hermès the court extended rights in a physical bag to NFT handbags, and in Juventus the court read a Class 9 registration to cover NFT cards. Both owners were famous, and a lesser brand may not get the same reading. A separate registration for virtual goods is therefore a safeguard, not a legal requirement, and no court has said otherwise.
Use. The United States treats three consecutive years of non-use as prima facie evidence of abandonment.72 The EU allows revocation after five years without genuine use.73 India allows removal after five years and three months. In each system, therefore, a brand that never sells a virtual version risks losing its virtual registration. None of the three has defined what counts as genuine use in a virtual world. In Yuga Labs the Ninth Circuit accepted that Yuga had used its marks in commerce through NFT sales; no other appellate court in the three systems has yet ruled on the point.
Confusion. US courts weigh multiple factors, and in Hermès survey evidence showed 18.7% net confusion. Yuga Labs shows that confusion must still be proved for NFTs. The EU asks whether the public is likely to confuse or associate the signs and goods. In Juventus, the court found confusion from identical signs and a link between the NFT cards and Class 9 goods. India asks the same question under Section 29, but no court has applied it to virtual goods. All three systems turn on whether virtual and physical goods are similar, and none has settled that.
Dilution. The United States protects famous marks under Section 43(c), and the Hermès jury found dilution. Jack Daniel’s holds that parody does not escape dilution liability when the mark is used as a source identifier. The EU protects marks with a reputation under Article 9(2)(c), without any need for similar goods. India protects reputed marks under Section 29(4). Reputation-based claims let famous brands skip classification questions, but smaller brands cannot use them. This is the main reason why well-known marks fare better than ordinary marks in virtual disputes.
Expression defences. The United States has the most developed rule. Rogers gave artists a threshold defence, and Jack Daniel’s removed it where the mark identifies source. The EU has no threshold test; expression enters through the “due cause” requirement and Article 14. One scholarly view holds that due cause is unclear and that creators cannot predict their liability.74 India has no express parody or artistic exception. Section 30 covers, among other uses, descriptive use, use that identifies goods as those of the proprietor, and use on accessory goods, and the Delhi High Court has handled expressive use case by case. No system gives clear protection to an NFT creator who brands a collection with another’s mark. The US rule is clear but shrinking, the EU rule is flexible but uncertain, and the Indian rule is thin.
Intermediary liability. In Tiffany, the Second Circuit held that general knowledge of infringement did not make eBay liable and that specific knowledge was needed. In the EU, L’Oréal held that a marketplace loses the hosting safe harbour if it plays an active role or knows of unlawful sales, and it allowed injunctions to prevent further infringements. The Digital Services Act now sets the hosting rules, including notice and action. In India, Section 79 of the Information Technology Act, 2000 gives intermediaries a safe harbour if they meet its conditions.75 The Delhi High Court has denied the safe harbour to a marketplace that played an active role.76 The EU rule is the most detailed. None of the three addresses a platform that mints NFTs, runs a virtual world or authenticates goods; Nike v. StockX raised that situation and settled before it was decided.
This comparison shows four points at which the systems fail the adequacy criteria set out in the conceptual framework: virtual use, similarity between physical and virtual goods, expressive use, and platform roles beyond hosting.
Gaps in current frameworks
1. Classification and registration. The Nice Classification now lists NFT-authenticated files in Class 9, and EUIPO and the USPTO require specific descriptions. The Indian Registry has issued no dedicated guidance on virtual goods, and virtual goods fit poorly in its classes. No treaty text deals with virtual goods. A brand owner cannot tell whether a registration for physical goods protects the virtual version. Criterion 1 fails, because owners and consumers cannot predict protection.
2. Use and non-use. The United States, the EU and India all allow cancellation for non-use after three to five years. None defines genuine use of a mark on virtual goods. A brand that registers for virtual goods and delays its launch cannot know what will count as use. Criterion 2 fails, because the protection of goodwill rests on an undefined test.
3. Similarity between virtual and physical goods. Confusion claims in all three systems depend on similar goods. The Court of Rome and the Hermès court found a link between physical and virtual goods for famous marks, but those results rest on fame and on their facts. No appellate court has set a general rule, and a brand without a reputation cannot rely on them. Criteria 1 and 2 fail for those brands.
4. Expressive use. Jack Daniel’s displaced the Rogers threshold where a mark identifies source, and the name of an NFT collection is likely to do so. The EU due cause test and Section 30 of the Indian Act make no express space for commentary and art. Creators cannot predict their liability, and brand owners cannot predict the defences they will face. All three regimes fail Criterion 3.
5. Platform roles. The EU and the United States have liability regimes for marketplaces that host listings. These do not cover platforms that mint NFTs, authenticate products, operate virtual worlds or sell as principals. Nike v. StockX involved several of these functions, and the case settled. The Indian safe harbour in Section 79 was not designed for these functions. Criterion 4 fails, because a platform’s responsibility depends on its label and not on its function.
6. Cross-border and anonymous enforcement. Registration and enforcement remain national matters, while virtual goods are sold across borders from anonymous wallets. TRIPS sets enforcement obligations but is silent on this context. Section 134 helps Indian litigants on forum, but it does not help them discover or serve an anonymous or foreign defendant. Criterion 5 fails.
The first three gaps therefore concern what trademark law protects, the fourth concerns the limits on that protection, and the fifth and sixth concern enforcement.
Recommendations
This paper recommends targeted changes to registry practice, statute and soft law. Each recommendation answers one gap, in the same order.
A. Classification guidance
The Indian Trade Marks Registry should issue examination guidance on virtual goods and NFTs. It should follow the current edition of the Nice Classification and require applicants to name the digital item, as EUIPO does. WIPO should add explanatory notes to Class 9 stating that a virtual good and its physical counterpart are separate goods for registration, so that brand owners know they must file for both.
B. A use standard for virtual settings
India, the EU and the United States should accept as use the sale or offer of a virtual good in an environment open to buyers in the territory. WIPO’s 2001 Joint Recommendation on the protection of marks on the Internet already uses a commercial-effect test for online use,77 and registries should extend it to virtual goods. Non-use periods should not run while the platform hosting the goods is closed.
C. A rebuttable similarity rule
A virtual good that depicts or replicates a physical good should be presumed similar to that good for the purposes of confusion analysis, and the defendant could rebut the presumption. The rule would give brands without fame the result that Hermès and Juventus reached for famous brands. It would belong in the Indian Act, in the EU examination guidelines and in US case law.
D. A defined space for expression
The law should state when a creator may refer to another’s mark without liability. The test should have three parts: the use must comment on or refer to the mark owner; the creator must not use the mark as the brand of its own collection; and the use must not suggest sponsorship. This follows Jack Daniel’s, which removed Rogers for source-identifying use, and it extends the referential use defences in Sections 30(1) and 30(2)(d) of the Indian Act and in Article 14 of the EUTMR to expressive works. The creator gains a clear route, and the brand owner keeps its claim against branding use.
E. Platform liability by function
Liability should follow what a platform does. A platform that only hosts listings should keep its safe harbour, subject to notice and action. A platform that mints tokens, authenticates goods or sells as principal should be treated as a seller. The EU already denies the safe harbour to hosts that play an active role, and this recommendation turns that principle into a standard. India should state it in guidance under Section 79 of the Information Technology Act, 2000, and the EU can apply it under the Digital Services Act.
F. Cross-border cooperation
The TRIPS Council or WIPO should publish guidance on three issues: first, when an online sale takes place in a territory; second, how anonymous wallet-holders may be summoned to court; and third, how registries and marketplace operators can communicate takedown notices. WIPO’s soft-law approach to well-known marks and online use provides a model. An amendment to TRIPS is not necessary: TRIPS only establishes minimum standards and says nothing about these issues, so guidance can fill the gap.
G. Limits
Two limitations apply. The recommendations rest on a limited body of case law, which Rothschild’s appeal in Hermès might change. They also put most pressure on India, which lacks any case law. For India, the quickest and least costly measures are Registry guidance, use criteria and the interpretation of Section 79, none of which requires legislation.
Conclusion
The purpose of this paper was to assess how trademark law protects trademark owners and consumers with respect to virtual goods. Trademark law largely suffices, except in six areas. Courts have extended rights over tangible products to virtual products, as Hermès and Juventus demonstrate, although both rulings rested on the fame of the mark and on their own facts. Separate registration in Class 9 is an important safeguard, although no ruling to date has held it necessary. In Hermès, despite the application of Rogers, liability followed because the jury found that Rothschild intended to mislead consumers. Jack Daniel’s has modified Rogers by removing it where a mark serves to identify source, and the name of an NFT collection often does so; expression is therefore confined to the confusion test, which reduces the room for art and commentary. Nike v. StockX was resolved before any judgment on its NFT-related claims, and the ruling against StockX concerned only its own sales of counterfeits. Of the existing rules on platforms, the EU active-role test for marketplaces is the most developed; the knowledge-based rule in Tiffany v. eBay and the safe harbour in Section 79 of the Information Technology Act, 2000 address hosting, and none of the three reaches platforms that mint tokens, authenticate goods or sell as principals. Virtual goods do not require sui generis protection, because the gaps are small and specific reforms can address them: Registry guidance on classification, a use standard for virtual environments, a rebuttable similarity rule, a defined space for expression, function-based liability for platforms, and international guidance on cross-border issues from the TRIPS Council or WIPO. India stands to benefit most from these measures. It has legislation but no case law, and Registry guidance, a use test and an interpretation of Section 79 require no new legislation. The paper is limited by a small and developing body of case law. Further research should follow Rothschild’s appeal in Hermès and the first Indian decision on virtual goods.
*****
Footnotes
1. Hermès Int’l v. Rothschild, 603 F. Supp. 3d 98 (S.D.N.Y. 2022).
2. Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989).
3. Jack Daniel’s Props., Inc. v. VIP Prods. LLC, 599 U.S. 140 (2023).
4. Nike, Inc. v. StockX LLC, No. 22-cv-983 (S.D.N.Y. filed Feb. 3, 2022).
5. See Trademark Infringement in the Metaverse and Virtual Environments, BNB Legal (Aug. 20, 2025), https://bnblegal.com/article/trademark-infringement-in-the-metaverse-and-virtual-environments/ (“India has not yet witnessed a landmark ruling on trademark infringement in the metaverse.”); Swarnadeep Das, Trademark Issues in Metaverse Commerce, Jus Corpus (Sept. 9, 2026), https://www.juscorpus.com/trademark-issues-in-metaverse-commerce/ (noting that India has no metaverse-specific trademark legislation).
6. Rocco Limongelli & Ludovica Sposini, The (Virtual) Battle for Intellectual Property Rights in the Metaverse: The Case of Copyright, Trademarks and the NFT Technology, 6 Metaverse 3056 (2025), https://doi.org/10.54517/m3056.
7. WooJung Jon & Sung-Pil Park, Comparative Analysis of Trademark Protection in the Metaverse and Registration of Virtual Goods and NFTs, 57 Comput. L. & Sec. Rev. 106137 (2025), https://doi.org/10.1016/j.clsr.2025.106137.
8. Ambarish Mangaraj, Navigating Trademark Rights in the NFT and Digital Goods Market, 6 Int’l J.L. Just. & Juris. 9 (2026), https://doi.org/10.22271/2790-0673.2026.v6.i6a.337.
9. Michael D. Murray, Trademarks, NFTs, and the Law of the Metaverse, 6 Ariz. L.J. Emerging Tech. (2023), https://doi.org/10.2458/azlawjet.5764.
10. Sakina Anwer & Tariq Hussain, Intellectual Property Rights in the Metaverse: Legal Framework and Future Prospects, 2 Contemp. J. Soc. Sci. Rev. 6 (2024), https://contemporaryjournal.com/index.php/14/article/view/1184.
11. Yvonne Draheim & Hendrik Schulze, Trademark Infringement Through NFTs: Intellectual Property Enforcement in the Virtual World, Hogan Lovells (Jan. 11, 2023), https://www.hoganlovells.com/en/publications/trademark-infringement-through-nfts-intellectual-property-enforcement-in-the-virtual-world.
12. Joshua Revilla & Sarah Bro, How the Metaverse and NFTs Might Impact IP Protection in 2023, World Trademark Rev. (Mar. 16, 2023), https://www.worldtrademarkreview.com/article/how-the-metaverse-and-nfts-might-impact-ip-protection-in-2023.
13. Jonathan Hyman & Eric Blosser, MetaBirkins Update: Is It Art or a Commercial Product?, The Recorder (Feb. 26, 2025), https://www.knobbe.com/wp-content/uploads/2025/03/The-Recorder-MetaBirkins-Update.pdf.
14. Aastha Gupta, Pixel Marks & Parallel Markets: Rethinking India’s Trademark Regime for Virtual Goods After Hermès v. Rothschild, Khurana & Khurana (Dec. 24, 2025), https://www.khuranaandkhurana.com/pixel-marks-parallel-markets-rethinking-india-s-trademark-regime-for-virtual-goods-after-herm%C3%A8s-v.
15. Das, supra note 5.
16. William M. Landes & Richard A. Posner, Trademark Law: An Economic Perspective, 30 J.L. & Econ. 265 (1987).
17. Frank I. Schechter, The Rational Basis of Trademark Protection, 40 Harv. L. Rev. 813 (1927).
18. Paris Convention for the Protection of Industrial Property art. 6bis, Mar. 20, 1883, as revised at Stockholm July 14, 1967, 21 U.S.T. 1583, 828 U.N.T.S. 305, https://www.wipo.int/treaties/en/ip/paris/.
19. Agreement on Trade-Related Aspects of Intellectual Property Rights art. 2, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 U.N.T.S. 299, https://www.wto.org/english/docs_e/legal_e/27-trips.pdf [hereinafter TRIPS]; see especially id. arts. 15(1), 16, 19(1), 41(1).
20. World Intell. Prop. Org., International Classification of Goods and Services for the Purposes of the Registration of Marks (Nice Classification) (12th ed. 2023), Class 9, https://www.wipo.int/classifications/nice. The current edition is the 13th (version 2026), in force since Jan. 1, 2026.
21. Stuart D. Levi, Mana Ghaemmaghami & Shannon N. Morgan, The Nice Classification Adds Language Related to NFTs, the Metaverse and Crypto Assets, Skadden, Arps, Slate, Meagher & Flom LLP (Feb. 1, 2023), https://www.skadden.com/insights/publications/2023/02/the-nice-classification-adds-language.
22. Matthew Lingard & Sarah Williams, EUIPO Guidance on Classifying Virtual Goods and NFTs, Walker Morris LLP (Dec. 1, 2022), https://www.walkermorris.co.uk/comment-opinion/euipo-guidance-on-classifying-virtual-goods-and-nfts/.
23. World Intell. Prop. Org., Joint Recommendation Concerning Provisions on the Protection of Well-Known Marks (1999).
24. The Trade Marks Act, No. 47 of 1999, § 2(1)(j), (zb), India Code (1999), https://ipindia.gov.in/tm-act-1999.
25. Gupta, supra note 14.
26. The Trade Marks Act, supra note 24, § 29(1), (4).
27. Id. §§ 2(1)(zg), 11.
28. Das, supra note 5 (“whether an avatar-based sale on a decentralised platform satisfies this phrase remains untested before an Indian court”).
29. The Trade Marks Act, supra note 24, § 30(1), (2)(a), (d).
30. Tata Sons Ltd. v. Greenpeace Int’l, 2011 SCC OnLine Del 466 (India), https://indiankanoon.org/doc/562656/.
31. The Trade Marks Act, supra note 24, § 47(1)(b).
32. Id. § 134.
33. Das, supra note 5; BNB Legal, supra note 5.
34. 15 U.S.C. §§ 1114(1), 1125(a)(1)(A), 1125(c), 1125(d), 1127.
35. Polaroid Corp. v. Polarad Elecs. Corp., 287 F.2d 492 (2d Cir. 1961).
36. AMF Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979).
37. Stuart D. Levi & Anita Oh, Jury Finds That “MetaBirkin” NFTs Infringed Hermès’ Trademark Rights, Skadden, Arps, Slate, Meagher & Flom LLP (Feb. 16, 2023), https://www.skadden.com/insights/publications/2023/02/jury-finds-that-metabirkin-nfts-infringed-hermes-trademark-rights.
38. Hermès Int’l v. Rothschild, No. 22-cv-384 (S.D.N.Y. filed Jan. 14, 2022), https://dockets.justia.com/docket/new-york/nysdce/1%3A2022cv00384/573363.
39. Hermès Int’l v. Rothschild, supra note 1.
40. Hermès Int’l v. Rothschild, 654 F. Supp. 3d 268 (S.D.N.Y. 2023) (opinion denying cross-motions for summary judgment, issued Feb. 2, 2023), https://law.justia.com/cases/federal/district-courts/new-york/nysdce/1%3A2022cv00384/573363/140/.
41. Hermès Successfully Defends Its Trademark in the Metaverse, Morrison & Foerster LLP (Feb. 9, 2023), https://www.mofo.com/resources/insights/230209-hermes-successfully-defends-trademark-metaverse.
42. Verdict, Hermès Int’l v. Rothschild, No. 22-cv-384 (S.D.N.Y. Feb. 8, 2023), ECF No. 144 (awarding $110,000 in net profits and $23,000 in statutory damages for cybersquatting); Final Judgment, Hermès Int’l v. Rothschild, No. 22-cv-384 (S.D.N.Y. Feb. 14, 2023), https://law.justia.com/cases/federal/district-courts/new-york/nysdce/1%3A2022cv00384/573363/145/.
43. Hermès Int’l v. Rothschild, No. 22-cv-384 (S.D.N.Y. June 23, 2023), https://law.justia.com/cases/federal/district-courts/new-york/nysdce/1%3A2022cv00384/573363/191/.
44. Alexandra D. Valenti & Timothy Keegan, MetaBirkins Post-Trial Ruling Clarifies Line Between Trademark Infringement and Free Expression and Grants Broad Injunctive Relief, Goodwin Procter LLP (June 28, 2023), https://www.goodwinlaw.com/en/insights/publications/2023/06/alerts-otherindustries-ip-metabirkins-post-trial-ruling-clarifies-line.
45. Brief for Defendant-Appellant, Hermès Int’l v. Rothschild, No. 23-1081 (2d Cir.), https://storage.courtlistener.com/recap/gov.uscourts.ca2.60801/gov.uscourts.ca2.60801.56.0.pdf; Brief for Plaintiffs-Appellees, Hermès Int’l v. Rothschild, No. 23-1081 (2d Cir.).
46. Hermès Int’l v. Rothschild, No. 23-1081 (2d Cir. argued Oct. 23, 2024).
47. Hyman & Blosser, supra note 13.
48. See Das, supra note 5 (reporting in September 2026 that “a final appellate ruling was still pending at the time of writing”).
49. Rogers, 875 F.2d at 999.
50. Jack Daniel’s, supra note 3.
51. Vans, Inc. v. MSCHF Prod. Studio, Inc., 88 F.4th 125 (2d Cir. 2023).
52. Yuga Labs, Inc. v. Ripps, 144 F.4th 1137 (9th Cir. 2025).
53. Nike, Inc. v. StockX LLC, supra note 4.
54. Nike v. StockX: A Timeline Behind the Trademark Lawsuit, The Fashion Law (Aug. 29, 2025), https://www.thefashionlaw.com/nike-v-stockx-a-timeline-behind-the-trademark-lawsuit/.
55. Stipulation of Dismissal, Nike, Inc. v. StockX LLC, No. 22-cv-983 (S.D.N.Y. Aug. 29, 2025).
56. Tiffany (NJ) Inc. v. eBay Inc., 600 F.3d 93 (2d Cir. 2010).
57. Yuga Labs, supra note 52.
58. Daniel Kuhn, Bored Ape Yacht Club Maker Yuga Labs Settles with ‘Expressive Appropriation Artist’ Ryder Ripps in Long-Running Trademark Case, The Block (Apr. 8, 2026), https://www.theblock.co/post/396788/bored-ape-yacht-club-yuga-labs-settles-appropriation-artist-ryder-ripps-trademark.
59. Daimler Benz Aktiengesellschaft v. Hybo Hindustan, AIR 1994 Delhi 239 (India), https://indiankanoon.org/doc/1460548/.
60. Tata Sons, supra note 30.
61. Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union Trade Mark, 2017 O.J. (L 154) 1, https://eur-lex.europa.eu/eli/reg/2017/1001/oj [hereinafter EUTMR].
62. Directive (EU) 2015/2436 of the European Parliament and of the Council of 16 Dec. 2015 to Approximate the Laws of the Member States Relating to Trade Marks, 2015 O.J. (L 336) 1, https://eur-lex.europa.eu/eli/dir/2015/2436/oj.
63. EUTMR, supra note 61, arts. 18(1), 58(1)(a).
64. Id. art. 9(2).
65. Id. art. 14(1).
66. Juventus F.C. S.p.A. v. Blockeras S.r.l., No. 32072/2022 (Trib. Roma July 20, 2022) (It.).
67. Draheim & Schulze, supra note 11; Lorenzo Battarino, Breaking: Italian Court Grants First Injunction Against NFTs Infringing Trademark Rights, IP in Italia (Nov. 10, 2022), https://www.ipinitalia.com/nft/breaking-italian-court-grants-first-injunction-against-nfts-infringing-trademark-rights/.
68. Case C-324/09, L’Oréal SA v. eBay Int’l AG, 2011 E.C.R. I-6011; Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 Oct. 2022 on a Single Market for Digital Services and Amending Directive 2000/31/EC (Digital Services Act), 2022 O.J. (L 277) 1, https://eur-lex.europa.eu/eli/reg/2022/2065/oj.
69. Revilla & Bro, supra note 12.
70. Das, supra note 5 (“The Indian Trade Marks Registry has not issued dedicated guidance on virtual goods”).
71. The Trade Marks Act, supra note 24, § 2(1)(j).
72. 15 U.S.C. § 1127 (defining “abandoned”).
73. EUTMR, supra note 61, arts. 18(1), 58(1)(a).
74. Mykola Antoniuk, Protection of Reputable Marks Beyond Confusion: Does “Due Cause” Help to Strike a Balance Between Trade Mark Proprietors and Content Creators?, Maastricht Univ. (Jan. 12, 2023), https://www.maastrichtuniversity.nl/blog/2023/01/protection-reputable-marks-beyond-confusion-does-%E2%80%9Cdue-cause%E2%80%9D-help-strike-balance.
75. The Information Technology Act, No. 21 of 2000, § 79, India Code (2000).
76. Christian Louboutin SAS v. Nakul Bajaj, 2018 SCC OnLine Del 12215 (India), https://www.itlaw.in/christian-louboutin-sas-versus-nakul-bajaj-ors-cs-comm-344-2018-intermediary-liability-ecommerce/.
77. World Intell. Prop. Org., Joint Recommendation Concerning Provisions on the Protection of Marks, and Other Industrial Property Rights in Signs, on the Internet arts. 2–3 (2001).