Technological advancements have always shaped the way we design, produce and consume fashion, and artificial intelligence (AI) is no different. From trend forecasting to supply chain optimisation, AI could prove transformative throughout the fashion industry and many brands have already provided insights into the AI initiatives they’re implementing. In keeping with the theme, I began my research into AI x fashion by asking ChatGPT, the AI-powered chatbot that has taken the world by storm, for some key topics covering the intersection between fashion and AI and have chosen to explore three of those topics.

1. The role of AI in fashion design

AI can aid fashion designers by predicting trends, anticipating popular products and forecasting consumer demand to inspire the design of items. Further, AI can design and generate visual images and videos to market and advertise products, as exemplified by Prada whose recent social media campaign featured AI-generated visuals of Prada’s most popular fragrances.

2. AI-driven personalisation in fashion

Using recommendation algorithms, virtual stylists recommend items to consumers who can in turn virtually try on those items before making a purchase. The growth of e-commerce has already caused a decrease in high street and shopping centre footfall, and such a personalised AI service could contribute to a further reduction.

My first thought was that AI could not replicate the human touches which some brands rely on as part of their identity, such as luxury brands. However, if AI can help provide a seamless sale process and instant customer service, both of which are expected when purchasing a luxury item of the highest quality, that could be powerful. LVMH, the French luxury conglomerate, announced a five-year partnership with Google Cloud Platform in 2021 to enable its brands to leverage new AI tools, including those which ensure “the highest standards of customer service and assistance”, evidencing luxury brands and AI as compatible.

3. Sustainable fashion, supply chain optimisation and AI

Joanne Crevoiserat, CEO of Tapestry Inc whose portfolio includes Coach, Kate Spade and Stuart Weizman, announced that Tapestry was “leverage[ing] new data analytic capabilities to optimise [its] product allocation processes, such as utilizing artificial intelligence to forecast customer demand, and better position inventory and stores”. This demonstrates AI’s ability to better allocate resources and contribute to more efficient supply chains which in turn can decrease waste. Matching supply with demand more accurately could help to reduce the 92m tonnes of textile waste which is produced every year, according to Earth.Org. AI could therefore help brands to meet their ESG goals, providing that the carbon footprint of the AI isn’t greater than that of the garments produced.

The challenges and limitations of AI within the fashion industry

Given the rapid evolution of AI, legal developments are required to provide clarity on numerous grey areas amidst the intertwining of AI and fashion. Legal nuance to cover such grey areas were required in the following two recent examples:

• Hermès v Rothschild – US trademark law had to adapt to the digital environment as Hermès challenged NFT creator, Mason Rothschild, following his creation of 100 faux fur ‘MetaBirkins’ (see my colleague Jessica’s article for more on this here). Ruling in favour of Hermès, the US court found that trademarks on real world items can also apply to digital goods and NFTs.
• Getty v Stability AI – this case centres on Stability AI’s unauthorised use of Getty’s images to train its AI programme without having negotiated a license for use of Getty’s content. This suggests that if fashion brands are to start creating and implementing their own AI models, a delicate road may lie ahead in selecting the assets they feed to AI models to train them in order to prevent any IP infringements or confusion as to the source of images.

Whilst AI is paving the way of a new era within fashion and the life cycle of fashion items, from design to consumption, legal clarification of previously unexplored territories is required. Given the rate at which AI popularity has exploded, the law is going to have to catch up quickly!