Deconstructing Artificial Authorship: Generative AI, Copyright Ownership, and the Protection of Textile Designs in the Global Fashion Market

Abstract

This article explores the legal implications of the use of generative artificial intelligence (AI) within the global fashion industry, particularly in relation to the copyright authorship and ownership of textile designs. Luxury houses and independent designers are more and more using computer models to create textile prints and avant-garde silhouettes, which is changing the way we think about copyright and human creativity. The paper aims to look at the issues around visual content made by artificial intelligence by studying the laws and court decisions in India, the U.S., and the E.U. This study uses a framework to analyse court cases, including Thaler v. Perlmutter, Naruto v. Slater, and Eastern Book Company v. D.B. Modak. It shows that present standards of originality are not enough to deal with questions of human-algorithmic co-creation and proposes a new legal allocation model to protect creative human labour.

Keywords

Generative Artificial Intelligence, Copyright Law, Textile Design Protection, Human Authorship, Originality Doctrine, Fashion Technology.

Introduction

The fashion and luxury industries are today experiencing a digital renaissance with the broad deployment of generative artificial intelligence systems. The aesthetic value of high-fashion garments has historically been strongly linked to human craftsmanship, pattern cutting, and manual graphic illustration. Today, deep learning architectures such as diffusion models and generative adversarial networks allow apparel conglomerates and individual design houses to generate high-fidelity fabric prints, complex embroidery schemas, and entire collections from textual instructions in real time. The integration of this technology exponentially increases supply-chain efficiency and opens up unprecedented avenues for visual experimentation, but also precipitates a profound structural crisis in global intellectual property law.

The core legal question algorithmic fashion design presents is the basic principle of human authorship. Copyright statutory frameworks around the world have been historically designed to safeguard the intellectual products of human minds during the era of industrial printing. If an autonomous software engine creates a brand new and highly complex textile pattern without human input as to the precise line, colour or shape, the resulting asset is in a legal grey zone.

The article discusses the systemic gap in current intellectual property regimes on whether statutory thresholds of copyright protection are satisfied by AI-assisted and AI-generated fashion prints. It examines how courts address the lack of direct human expression in automated outputs in balancing corporate assets against public access.

The paper argues that the outright denial of copyright protection to AI-generated fashion designs creates a dangerous vacuum in the public domain that rewards large-scale corporate copyists, while full copyright over completely automated works is granted unconditionally, diluting the core purpose of creative protection. In sum, this article shows that a progressive legal framework must recognise an intermediate model of “joint human-algorithmic curation” to sustain equity in the global fashion economy.

Background / Conceptual Framework

To properly evaluate the disruptive potential of generative technology in the fashion sector, it is necessary to set the conceptual boundaries of copyright authorship and the doctrine of originality. Abstract ideas, functional concepts, or stylistic themes are not protected; only the original expression of an idea fixed in a tangible medium. This is the philosophy of copyright. In the fashion industry, this distinction is applied through the useful articles doctrine and the separability principle. Apparel serves a functional utilitarian purpose of covering the human body. The structural shape of a garment is generally excepted from copyright protection. Instead, copyright protection is directed to the conceptually separable surface ornamentation, graphic patterns, and new textile designs applied to the fabric.

To be protected, a textile design must be sufficiently original. The rules for deciding what is original and can be copyrighted are very different in countries.

There are three ways to think about this: the ‘Sweat of the Brow’ idea, which’s about how much work someone puts into something, and the ‘Modicum of Creativity’ idea, which is about whether something is truly creative.

– In India, the copyright law says that artistic works are protected if they are original, which means they have to show some skill and judgment and not just be copied from someone. The Supreme Court of India has said that this means the artistic works have to have a bit of creative choice and cannot be just mechanical or a copy. Luxury houses, independent designers, and artificial intelligence are changing the way we think about luxury houses and independent designers and artificial intelligence.

– In the United States, under Section 102(a) of the Copyright Act of 1976, a minimum degree of creativity is required for copyright law, and the work must be fixed in a digital medium. This is what the Copyright Act of 1976 says about copyright law in the United States.

The conceptual friction arises when these statutory frameworks are applied to generative artificial intelligence. An AI model has no intention, no conscious judgement, no subjective creative expression. It works by analysing mathematical relationships between vast amounts of man-made clothing and producing pixels based on probability. This algorithmic structure challenges the traditional legal idea of the creator, which forms the baseline for the legal and judicial analysis to be made.

Legal Analysis

The legal crisis over generative artificial intelligence in fashion comes down to an open question of statutory interpretation: can the prompt-driven Is the orchestration of an algorithm be a legally protectable creative effort? In traditional fashion production, the illustrator working with digital software such as Adobe Illustrator is the obvious author because the software is a purely mechanical tool that translates human muscle movement and aesthetic decisions directly into pixels.

This change affects the legal relationship between the human operator and the final artistic output. In keeping with the current United States copyright regime, the US Copyright Office has taken a conservative position that if the traditional elements of authorship in a work are generated by an algorithm, the work is an algorithm and cannot be registered, because it’s an algorithm, not a human, who created it.” The Office says that the text prompts are closer to the normal commercial instructions given to an artist who has been commissioned than to direct creative expression. If a fashion design house uses automated software to design a seasonal capsule collection, those patterns are instantly vulnerable to competitors. Unregistered AI prints can be copied by mass-market retailers to produce identical garments without fear of liability for copyright infringement because the source designs remain permanently in the public domain.

The Court of Justice of the European Union (CJEU) has a harmonised standard of originality in the European Union’s legal field. A work is original if it is the author’s own intellectual creation. In order to meet this requirement, the author should be able to make free and creative choices that leave a personal stamp on the final work. European scholars argue that although the initial choice of prompts and iterative filtering of generated variations are human choices, the uncontrollable randomness of diffusion models prevents the output from truly reflecting a human personal stamp. Such a rigid approach may put European luxury houses at a disadvantage compared to regions with more flexible copyright registration systems.

At the same time, the Indian statutory regime takes a different view under Section 2(d)(vi) of the Copyright Act, 1957. (in relation to any literary, dramatic, musical or artistic work which is computer-generated) the person who causes the work to be created “ The statute addresses computer-generated works by defining the author as ” This language seems to create a loophole for AI fashion design since a brand owner could claim the person who “causes the work to be created” is the model developer or the designer who inputs the text prompts.

However, this provision is yet to be tested in Indian courts against modern generative AI models that work with independent autonomy. The use of Section 2(d)(vi) without well-defined boundaries would be a grave legal danger as it could allow corporate bodies to obtain sweeping copyright monopolies in millions of automatically generated design versions, thus practically barring independent human designers from using common cultural symbols and geometric designs.

The same statutory ambiguity applies to the governance of platforms and online marketplaces. When an algorithmic system repeatedly digs through digital databases, it often uses copyrighted textile designs owned by independent artists without their permission to train its neural networks. This creates a dual risk of liability: liability for the platform for the developers of the technology that hosts the infringing datasets, and liability for downstream infringement on the part of the fashion brands selling garments with AI-generated prints that include fragments of protected works. The world of fashion lies in between a tight human-only registration system that denies the commercial value of automated apparel and an overly broad corporate model that risks monopolising the fundamentals of creative design.

Case Law Discussion

Case 1 : Thaler v. Perlmutter (D.D.C., 2023)

The precise legal question was whether a fully autonomous computer algorithm could legally be considered an “author” under the United States Copyright Act, and whether a work generated entirely by an AI system without human intervention could be given statutory registration.

The dispute concerned the language of Section 102 of the Copyright Act of 1976 and centuries of common law precedent of copyright protection requiring human authorship as an unalterable statutory prerequisite.

The court did not agree with the claimant that copyright law should change to encourage creativity. The court said that the law has always meant that only human beings can be authors. The court also said that the purpose of copyright law is to help people make money from their work. If machines could get copyrights, that would go against what the law’s meant to do.

The District Court decided that human authors are necessary for something to be copyrighted. This means that fashion designs made completely by machines are still not protected by US law.

Case 2: Naruto v. Slater (9th Circuit, 2018)

The question was whether a monkey could sue for copyright infringement because it took a selfie with a camera that nobody was using. The Court had to decide if animals can go to court to protect their copyrights.

The rule is that only human beings can enforce their intellectual property rights under the Copyright Act. This is a principle of law that can only be changed by Congress.

The Ninth Circuit decided that even though the monkeys’ photos were original and creative, the fact that a human being did not take them was the point. The court said that only human beings can be considered authors under the law unless the law is changed to say so.

In the end, the Ninth Circuit ruled that animals cannot hold or enforce copyrights. This means that if an animal creates something, it will be very hard for that creation to be protected by copyright law.

Case 3: Eastern Book Company v. D.B. Modak (Supreme Court of India 2008)

The main issue in Eastern Book Company v. D.B. Modak was about what’s needed for a book that is compiled or changed by a machine to be protected by copyright in India.

The Supreme Court of India looked at the idea of the labour doctrine, which is used in law, and compared it to the idea of needing a level of creativity for something to be protected by copyright in India.

The court decided that just putting in a lot of work is not enough to make something eligible for copyright protection. The court said that for something to be original, it needs to have a mix of skill and judgment, and this shows a level of creativity.

So in Eastern Book Company v. D.B. Modak, the court said that if a person uses their judgment and selection when creating something with a machine, then that thing can be protected by copyright in India.

Critical Analysis/ Findings

When we look at copyright laws in countries, we can see that there is a gap in the laws that govern the technology used in the fashion industry. The fashion industry is facing a problem. The old way of thinking about copyright, which says something is either made by a human or made by a machine, is not accurate anymore. The fashion industry is what we are talking about here.

The fashion industry uses machines to create clothes. Humans are also involved in the process. Humans. Improve the work before the final product is made. The fashion industry uses a combination of machine input and human judgment.

The court’s decision in Eastern Book Company v. D.B. Modak helps us see how important human judgment and skill are in creating something that can be protected by copyright. The fashion industry needs to know this. The fashion industry is what we are talking about. The fashion industry is facing a challenge because of the gap in copyright laws. The fashion industry needs to understand how copyright laws apply to their work.

A binary copyright system creates one clear advantage and one disadvantage for various parties involved in the creative process. Large fast-fashion companies currently benefit substantially from having no copyright protection; they regularly use automated web scraping bots to copy AI-generated prints from independent designers and subsequently manufacture products with those prints without facing any financial consequences for doing so. Conversely, independent designers and new entrants into the fashion technology ecosystem have no legal recourse to protect their technology-influenced collections from being unlawfully copied and reproduced digitally.

 The existing legal framework also fails to address the systemic issue of third-party data scraping during the model training phase. Current legal frameworks allow large technology platforms to create generative models that were trained with unlawfully obtained datasets, since there is broad parity under fair use.

Right now, the law is designed to allow businesses to utilize the artistic labour of others (humans) without permission in order to assist in building their commercial automation tools.

The challenge we must meet is to develop an international copyright system that does not adhere to the rigid structure of the current binary system; instead, we need a new intermediate structure in which there will be a separate type of copyright for “AI-Assisted Human Curation”. In this model the copyright will be held for the software engine (i.e., the generative algorithm), not just by the software developer, but also by the human creator of the image/print if that creator can demonstrate adequate and extensive manual post generation modifications and extensive iterative refinement of the image/print through careful manipulation of the prompts used in order to produce the images/prints. This model will provide businesses with protection for the money they have invested into developing new digital fashion products, while still ensuring that the human contribution continues to remain the basis of copyright protection around the world.

Conclusion

This article illustrates that with the use of generative artificial intelligence technology, there presents an entirely new structural challenge to the foundational principles of traditional copyright law, particularly within fashion. The analysis done in this study indicates that the original standards of human authorship and originality are incapable of adequately addressing the issues associated with the increasing amounts of human/algorithm hybrid co-creation. If we were to provide no protection to technology-based textile prints, then there would be an unprotected public domain for artistic works that are subject to possible exploitation by bad-acting businesses who copy original designs without giving credit to the original human creators. On the other hand, if we provide absolute exclusive rights to businesses that utilise generative artificial intelligence (i.e., using artificial intelligence as a tool, but producing the artwork or print themselves), we are putting the market at risk of being flooded with solely automated/algorithmically generated designs

In order to achieve a stable market, it will be necessary for global copyright systems to modify the way statutes are interpreted. Legislative bodies should create one set of clear, standardized guidelines that define both human prompt engineering and iterative post-generative changes as a legitimate form of creativity. Given the current trends in the fashion business moving into a virtual space (and likewise all creative industries), updating the standards for Intellectual Property (IP) protections needs to be done so as to allow copyright law to optimally protect the creativity of humans, including encouraging technology development.

Reference(S):

1. Abbott, R., & Rothman, E. (2024). A utilitarian approach to copyright law and generative artificial intelligence. In R. A. Posner (Ed.), Research handbook on law and utilitarianism (pp. 349–367). Edward Elgar.
2. Alam, I. (2015). Provision of copyright for commercial legal database with special reference to SCC. In Transforming dimension of IPR: Challenges for new age libraries (p. 257).
3. Babie, P. T. (2018). The monkey selfies: Reflections on copyright in photographs of animals. UC Davis Law Review Online, 52, 103.
4. Candelon, F., & Reeves, M. (Eds.). (2022). The rise of AI-powered companies. Walter de Gruyter.
5. Cupp, R. L., Jr. (2020). Considering the private animal and damages. Washington University Law Review, 98, 1313.
6. Dhananjaya, C. (2025). Challenges and solutions for copyright infringement in the digital age. LawFoyer International Journal of Doctrinal Legal Research, 3, 1670.
7. Gonzalez, N. C. (2021). Legal personhood and animals. University of Central Florida Department of Legal Studies Law Journal, 4, 163.
8. Hooker, M. P. (2020). Naruto v. Slater: One small snap for a monkey, one giant lawsuit for animal-kind. Wake Forest Law Review Online, 10, 15.
9. Jain, S. (2025). Impact of artificial intelligence on intellectual property law in India. Supremo Amicus, 38.
10. Kiliç, S., & Ediş, L. B. (2026). The digital future of the fashion industry: Slow fashion and artificial intelligence approach to sustainability. In Rethinking fashion and consumer responsibility for decent work and social justice (pp. 273–324). IGI Global.
11. Lu, S. (2023). Deepfakes under copyright law—A necessary legal innovation. Southern Illinois University Law Journal, 48, 517.
12. Raza, A. (2021). Theoretical underpinnings of copyright and design laws: Decisions of the Supreme Court of India. Aqa Raza, 26(4), 220–234.
13. Ringer, B. (1976). First thoughts on the Copyright Act of 1976. New York Law School Law Review, 22, 477.
14. Salama, S. J. (2019). Is it so easy a monkey can do it: Joint works and the unintended collaborator. Hofstra Law Review, 48, 215.
15. Schneider, D. (2018). Why a monkey’s action of taking a selfie should expand the definition of an author in the Copyright Act. Touro Law Review, 34, 1349.
16. Singh, P. M. (2020). Evolution of copyright law—The Indian journey. Indian Journal of Law and Technology, 16, 38.
17. Sontakke, A., & Bhatt, H. (2012). Scope of rights of broadcasting organisations under Copyright Act, 1957. Rajiv Gandhi National Law University Student Law Review, 1, 102–121.
18. Thaichon, P., & Quach, S. (Eds.). (2023). Artificial intelligence for marketing management (Vol. 4). Routledge.
19. Zrno, G. (2023). Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union trade mark. Hum: Časopis Filozofskog fakulteta Sveučilišta u Mostaru, 18(30), 54–75.