• Home
  • Education
  • News
  • Fashion's AI Dilemma: Innovation or Imitation?

Fashion's AI Dilemma: Innovation or Imitation?

By Guest Contributor

loading...

Scroll down to read more
News
AI image that illustrates plagiarism in fashion. Credits: generated by FashionUnited with an AI tool. Credit: FashionUnited

In the near future, generative AI could add up to US$275 billion to the apparel, fashion, and luxury sectors’ operating profits, according to McKinsey. While this transformative technology promises to revolutionise design, production, and even the way we experience fashion, there is a crucial need to navigate the ethical and legal complexities that AI presents in order to truly realise its potential. What are the implications that we need to be aware of?

ABOUT
Written by Erin Cho, Dean and Limin Professor in Integrated Strategies and Leadership in Fashion, School of Fashion and Textiles at The Hong Kong Polytechnic University.

One of the most pressing concerns is the emergence of deepfakes, or hyper-realistic AI-generated images and videos that can be used for both harmless entertainment and malicious purposes. While Katy Perry's playful use of a deepfake to create a digital appearance at the Met Gala highlights the potential for creative expression, the technology can also be exploited to spread misinformation and damage brand reputation. Cassey Ho, founder of Popflex, faced this issue when a deepfake video featuring her body but a different face was used to promote a counterfeit version of her skort (skirt with shorts underneath) on Amazon.

This incident underscores the potential for deepfakes to erode consumer trust, violate personal rights, and create a climate of uncertainty surrounding online content. Therefore, it necessitates a deeper understanding by professionals in the fashion industry in terms of AI-generated content and how to distinguish it from authentic creations. Moreover, models and influencers need to be proactive in securing their digital identity and protecting their online presence from manipulation through deepfakes. Understanding the technology behind them and taking steps to secure their social media accounts and online profiles would be a necessary first step.

Beyond deepfakes, the rise of AI-powered design tools raises questions about the ownership of creative work. Shein has recently expanded partnership with NTX Group, a textile innovation and solutions company to implement AI into its design processes, enabling it to bring new items to market within a week. Yet, the fast fashion giant has also been accused of leveraging AI to track online trends and copy designs from artists and established brands without proper attribution or compensation. This practice raises concerns about intellectual property rights and the potential for AI to exacerbate existing issues of plagiarism and copyright infringement in the fashion industry.

To address these concerns, organisations like the Council of Fashion Designers of America (CFDA) are partnering with companies like Raive, which builds IP-safe AI models, to foster brand protection within the fashion industry. Raive's model is designed to attribute the use of specific intellectual property and provide royalties to official licensed IP holders, offering a potential solution to the concern that AI systems will utilise protected works without credit or compensation. This approach also allows brands to isolate their intellectual property from the base model, ensuring that AI-generated designs are based solely on their own data.

Now more than ever, designers should be aware of the legal framework for protecting their designs from unauthorised use. While registering patents is one possible method, thanks to more advanced technologies, designers can now explore digital watermarks that can help track their work’s origin and prove ownership in case of infringement. It is also prime time for fashion professionals globally to unite and set industry standards on the ethical and responsible use of AI in design, and to advocate for stronger international laws and regulations that protect intellectual property rights in the digital age.

The issue of transparency surrounding AI-generated content remains a major challenge. While platforms like TikTok are adding "AI-generated" labels to content created with AI tools, inconsistencies in labelling and the lack of clear guidelines raise concerns among creatives who fear their work being misidentified. OpenAI's integration of C2PA metadata into its AI models represents a significant step towards addressing this issue. This metadata provides a traceable record of how AI-generated content was created, making it easier to identify and attribute work to the appropriate creator and enhance trust in AI-generated content.

The future of AI in fashion hinges on a delicate balance between innovation and ethical considerations. While AI offers exciting opportunities for creative expression, it's crucial for creatives, brands, and consumers to be aware of the potential pitfalls related to authenticity, ownership, and transparency. By addressing these concerns through responsible development, clear labelling, and robust IP protection measures, we can ensure that AI empowers creativity while upholding ethical standards and protecting the rights of creators.

AI
Education News
Hong Kong Polytechnic University
School of Fashion and Textiles