The fashion industry has been an early adopter of artificial intelligence, leveraging it to enhance customer experiences and streamline operations. In recent years, AI's impact on fashion has increased, bringing innovative applications alongside new legal complexities.
While most legal challenges faced by fashion brands using AI - such as those related to intellectual property and privacy - are addressed under existing laws, the EU’s upcoming AI Act introduces additional regulatory implications. Fashion provides a compelling example of how the AI Act might influence industries typically viewed as “low risk” under the Regulation’s risk-based framework.
This article examines the main AI applications in fashion, the legal challenges they face under existing regulations, and how the EU AI Act may further shape compliance obligations and practices for fashion brands.
AI in fashion: a variety of use cases
The fashion industry has embraced AI for several years, integrating it into multiple areas to boost productivity, creativity, and customer engagement. Among the most prominent use cases are:
- Enhanced shopping experience - as in many industries, the place of AI in customer service has grown substantially in recent years. AI enables fashion brands to offer instant customer support, reducing response times and enhancing service quality. It is also widely used for personalised recommendations, leveraging a broader array of input data beyond simple purchase history or preferences. This wider scope of consumer-provided input data, including potential biometric information, enhances the customisation of products to meet individual needs. This also results in AI-driven virtual try-ons, allowing shoppers to preview clothing on themselves through augmented reality (AR) applications, creating a more immersive and tailored shopping experience.
- Virtual models and imagery - AI, and more specifically generative AI, has a particularly significant impact in the fashion industry by enabling brands to create high-quality visuals while minimising the costs and environmental impact of traditional photo shoots. By training AI on their existing image datasets, brands can also seamlessly integrate new products into established visual themes, allowing for the effortless addition of new items to collections without the need for additional photo shoots. This trend has even extended to the use of virtual models and influencers. Some brands now embrace having their products showcased by entirely virtual personas.
- Integration of AI in the design process - AI is now also integrated into the core of the fashion industry’s value chain: the design and creation process. Advanced algorithms enable fashion brands to analyse large-scale, real-time trend data, allowing them to anticipate future consumer preferences more accurately and quickly. Generative AI takes this further by empowering designers to create new patterns and designs inspired by current online trends and past collections. This application of AI is said to be particularly advantageous for the fast fashion sector, where rapid adaptation to emerging trends is essential for success.
Current legal challenges faced by the use of AI in the fashion industry
The use of AI in the fashion industry brings significant legal challenges, starting with intellectual property concerns. AI’s ability to replicate designs and analyse existing collections raises potential counterfeiting risks. Virtual models and AI-generated designs may inadvertently mimic styles or resemble proprietary brand elements, also leading to possible legal disputes over intellectual property rights.
While the EU’s AI Act does not directly address these intellectual property issues, its article 53 does require providers of general-purpose AI models to implement state-of-the-art technology to identify and respect any data mining opt-out expressed in accordance with the provisions of the Copyright Directive. This provision of the AI Act gives fashion brands an additional reason to explicitly reserve their rights in a machine-readable format if they wish to prevent their collections and images from being used to train generative AI model.
Privacy compliance is another critical issue for the fashion industry as many AI-driven retail strategies rely on extensive processing of personal data. Brands utilising data for personalisation and virtual try-on experiences must navigate GDPR requirements, particularly when sensitive data, such as biometric information, is involved.
These legal risks exist not only when leveraging large external datasets but also when building or integrating AI models using existing brand data. Contracts with models and designers may notably need revision, as these agreements often do not account for the training and use of AI technologies.
The fashion industry is currently working to identify and address these complex legal risks as AI becomes central to their business strategies. With the introduction of the EU AI Act, the industry will face an added layer of compliance obligations.
EU AI Act: increased transparency requirements for the fashion industry
The EU AI Act introduces a risk-based framework for AI, primarily addressing high-risk systems. Most of the general AI use cases in fashion mentioned above - like virtual try-ons, digital imagery, and AI-assisted design - are unlikely to fall under the high-risk or prohibited categories of the regulation. Despite the extensive use of AI in fashion, most brands will also not qualify as “providers of AI systems” under the AI Act. However, they will still be classified as “deployers” of AI, which brings specific responsibilities. All deployers of AI systems have media literacy responsibilities but where the AI is high-risk or of a type specified in Article 50, there will also be transparency and accountability obligations (see here for more on deployer responsibilities).
Article 50 of the AI Act mandates that brands using AI clearly disclose AI-generated content to consumers. The Act’s definition of “deepfake” is broad, covering any AI-generated images that resemble real people, objects, places, or events that could be mistaken for authentic. This likely applies to AI-generated visuals used in place of traditional photo shoots, making it essential for fashion brands to integrate these transparency requirements into their marketing strategies and visuals.
Further, the Act imposes additional transparency requirements for deployers of AI systems that involve “biometric categorisation,” potentially relevant to some virtual try-on tools used by fashion brands. This section underscores the need for AI deployers to align with GDPR and existing data privacy regulations, emphasising data protection compliance in these AI-driven experiences. Meeting these enhanced transparency standards will require fashion brands to rethink how they communicate AI usage to consumers and integrate transparency effectively.
Since many fashion brands will rely on third-party AI providers to implement their AI capabilities, updating vendor compliance processes will also be critical. Brands will need to address these transparency points with their AI providers, particularly around the data and bias used to train the AI.
The Act’s transparency requirements applicable to AI providers will support this need, but fashion brands should proactively implement procedures to request and review technical documentation. Understanding the specifics of this documentation will ensure they are fully informed and able to communicate transparently about the AI systems they deploy.
AI Act as an opportunity for the fashion industry to rethink its AI compliance strategy
While the EU AI Act does not drastically alter existing compliance obligations and legal challenges for the majority of fashion brands using AI, it sets a new standard for transparency and accountability which brands may find useful to adopt even if not necessarily required to do so. The regulatory shift can be seen as offering fashion brands a valuable opportunity to adopt ethical and transparent AI practices, ultimately building greater consumer trust.
As consumers become more aware of AI’s role in their interactions with their favourite brands, their expectations for transparency will likely extend beyond simple disclaimers. Fashion brands should be prepared to meet these expectations by enhancing transparency policies and offering clear information on how AI is deployed in the whole customer experience.
For the fashion industry, the EU AI Act provides a framework for responsible AI compliance that aligns with broader trends in ethical and sustainable business practices. Embracing these changes can not only help mitigate legal risks but also assure customers of fair and transparent AI use throughout their engagement with the brand.
