EVERY week, we read about authors and artists in copyright lawsuits against companies behind generative artificial intelligence (GenAI) applications such as ChatGPT and Midjourney. Presently, there are about 50 ongoing lawsuits globally claiming that the use of copyrighted works such as literary texts, paintings and photographs to train GenAI are illegal, or that the outputs produced are substantially similar to the originals, and hence have a substitutive impact on these works.
The clarion call by the creators urges for stronger copyright enforcement. The fashion industry, however, presents a striking paradox: This global powerhouse, with a projected revenue of US$1.7 trillion in 2025, has long thrived in the face of weaker copyright laws.
Copyright’s “negative space” is an area where creation and innovation can thrive without significant protection from the law. In this negative space, creators are not exclusively motivated or incentivised by the prospect of intellectual property (IP) protection.
Now, two technological forces – the proliferation of social media and the rise of GenAI applications – are set to enlarge this space, creating both unprecedented opportunities and challenges.
This environment is perhaps a low-IP zone rather than a no-IP zone. Around the world, trademark laws provide significant legal protection to insignias that designate source of goods. Patents offer a valuable monopoly over novel and useful processes, manufacture or fabric; and registered designs grant protection for new, non-obvious, ornamental and non-functional designs.
However, copyright law offers the weakest protection of them all. Yet, this weakness is a feature that drives the industry forward.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
Copying in the fashion industry
Much of fashion borrows from the past. Some would say “copy”.
Fast fashion – exemplified by retailers such as Shein, Zara, H&M, Fashion Nova and Boohoo – is built on a business model of translating ideas from the runway shows of Paris, Milan, New York and London or celebrity culture into affordable garments at breakneck speed.
Fast fashion offers “knock-offs”, as opposed to “counterfeits”. A knock-off is an imitation of the original, with subtle differences copying the protected aspects of the original; it does not replicate the logo of the original and usually does not run afoul of IP laws.
A counterfeit, in contrast, is virtually indistinguishable from the original, intentionally tries to pass off as the original – right down to the logo – and would clearly infringe IP laws as well as attract criminal sanctions.
This legal room for knock-offs benefits the entire fashion industry through the joint operation of anchoring and induced obsolescence. Fashion copying is instrumental in creating trends (anchoring a specific look in the public consciousness), and saturating the market with similar goods until they become unstylish (induced obsolescence).
“The law protects only the expression of the idea, not the idea itself.”
This drives the consumers’ desire for new, emerging styles, which will, in turn, be copied and become unstylish in a never-ending cycle.
Many luxury fashion brands compete to set a trend – for instance, to be the first to launch the “It” bag of the season or a particular print on a garment. In a low-copyright environment, first-mover advantages and network effects usually explain how originators of trends continue to thrive.
A first-mover advantage exists when a creator or innovator can create enough revenue from the introduction of its product that it is not effectively harmed by later copyists – there is value to being the “genuine article”.
For consumers, there is enduring value in owning the “genuine article”, and the fashion cycle itself embodies a powerful set of first-mover advantages for style originators.
The legal foundation for this is the idea-expression dichotomy. Copyright tends to protect artistic works – including pictorial, graphic and sculptural works – but only their aesthetic or expressive features, and not their mechanical or utilitarian aspects.
The law protects only the expression of the idea, not the idea itself. For instance, Versace’s iconic gold Barocco print has been widely copied (not identically) in fast fashion knock-offs, as the common idea of a “gold baroque floral design” is not protected by copyright; only Versace’s specific expression of it is.
This significant negative space in the fashion industry has allowed fast fashion and small businesses to thrive. Because of this rapidly shortening of fashion cycles, knock-offs can benefit designers by inducing faster turnover, additional sales and a constant demand for new products.
How technology enlarges negative space
Social media is both a boon and bane for luxury fashion brands. Images and videos of new designs unveiled at the seasonal runway shows are widely circulated on Instagram and X just minutes after, or even streamed simultaneously, and often by celebrity influencers.
This massively increases a brand’s reach to consumers and fans, and can significantly boost sales. McKinsey & Co reported that 38 per cent of brand discovery today happens on social media.
However, this also gives fast fashion copyists immediate access to the designs, allowing them to start production and make them available for purchase within weeks of the shows, while the actual authentic collections are launched only six to eight months later.
While the social media posts could be infringing copyright in the expressive features of the fashion designs, the unauthorised reposting on social media is usually tolerated (in many situations, even necessary for the luxury fashion brands).
What is potentially damaging for these brands are the click-through functions – including ads – on social media, which allow consumers to immediately buy knock-off designs and even counterfeits.
The strategic opportunity here lies in leveraging the same platforms for brand advantage. Brands may use social media to draw attention to these duplicitous knock-offs (such as engaging with communities like Diet Prada that call out copies) and encourage consumers to go for the real thing.
Luxury brands can turn cheap knock-offs into their commercial advantage through more creative social media content that contrasts the products. The “copying” – euphemistically called “reinterpretation” or “homage” – can also occur between the high fashion brands. Calling out unattributed referencing on social media can result in drawing positive attention to the original designer and brand.
In addition, the rapid adoption of GenAI has introduced unprecedented challenges to copyright law, particularly as AI-generated works become increasingly prevalent in creative industries.
Human authorship remains a foundational concept in copyright law, and its application to AI-generated works raises some of the most important legal questions of the digital age.
One such challenge is determining whether AI-generated outputs are eligible for copyright protection. Recent decisions from the US Copyright Office and Beijing Internet Court, while seemingly divergent, are cohering into a consensus: For an AI-generated work to be copyrightable, there must be significant, creative human contribution and control over the AI-assisted creation process.
“We are witnessing the democratisation of fashion – the result of more individuals being able to utilise technology to build on existing works to create new creations and innovations at a faster pace. ”
The March 2025 decision of the Chinese Zhangjiagang Court in the “jelly butterfly chairs” case aligns with this. The plaintiff’s AI-generated pictures on Midjourney lacked enough original authorship to be copyrightable, as she could not provide adequate documentation to prove that she had made aesthetic choices and personal judgement during the generative process.
The business implication is clear: Inputting simple prompts is insufficient to make a user the author of the AI-generated output. Even if the human individual who provides the prompts is deemed the “author” of the resulting AI-assisted output, the idea-expression dichotomy will nonetheless operate to render ideas and styles unprotectable under copyright law.
This high threshold for copyrightability enlarges copyright’s negative space in fashion, but it is not necessarily a bad thing. Anyone wanting to use a GenAI application to create “gold baroque floral prints” can generate countless variations.
Each output is unlikely to attract copyrightability; if it does, it would be entitled only to thin copyright protection. Practically, this means anyone can freely generate and commercially exploit the use of gold baroque floral prints on dresses.
Even faster fashion lies ahead
A defining characteristic of the fashion industry today is choice, with a legion of most non-infringing substitutes for nearly every trend. According to McKinsey, Gen Z has popularised “dupes”, with the #dupe hashtag on TikTok attracting six billion views. In addition, one-third of US adults have said that they intentionally bought a knock-off of a premium or luxury product.
Whoever comes out on top depends on one’s ability to originate a trend, engender desire, and build customer loyalty.
As technology develops at a breathtaking pace and more GenAI systems become freely accessible, the traditional mode of producing, disseminating and licensing literary, dramatic, musical and artistic works will also have to evolve.
We are witnessing the democratisation of fashion – the result of more individuals being able to utilise technology to build on existing works to create new creations and innovations at a faster pace.
This could be viewed as a welcome renaissance in which AI has enabled anyone to be a fashion designer. In his interviews with artists between 2017 and 2020 – unlike the current backlash – researcher Ahmed Elgammal discovered that many artists who worked with AI found that it gave them sparks of new ideas, directions and ways to create their art.
Perhaps the future is not one of the decline of the arts, but the rise of creativity for more of us.
The writer is a professor and co-director of the Centre for Technology, Robotics, Artificial Intelligence & the Law at the National University of Singapore. He is also the lead editor of the upcoming book Fashion and Intellectual Property.
