How the fast fashion industry could be gambling with your trust when it comes to implementing AI.
The insatiable appetite for data might be a liability for the fast fashion industry if they are not more transparent. Francesca Mazzi, a research fellow in AI and sustainable development at Oxford, provides a warning for the industry. The Said Business School professor listed several challenges facing the industry in her 2020 article titled “Artificial Intelligence and the challenges of the fashion industries”. Mazzi presented her thoughts on how the fashion industry could apply AI to their biggest challenges: inaccurate trend forecasting and optimizing product inventory. Mazzi suggested that AI will be used to identify design trends during the design phase, which could reduce product forecast errors by 50% and reduce overall inventory requirements by 20-50%. Mazzi also pointed out that AI could play a role in sustainability as well, stating that high accuracy forecasts would also curb over-production and reduce unnecessary environmental impact. Given the industry's gaffes in sustainability and recent trust issues there is certainly a lot to be gained and potentially lost with the implementation of these new technologies.
There are many things that will have to go right if the implementation of AI in fast fashion is going to have a positive effect, and the industry will need to walk an ethical tightrope to maintain trust with its customers. To drive accurate use of AI applications in the industry, large amounts of consumer data will be needed to predict the fashion trends that will be most likely to become best-selling product lines. The data required to fuel trend forecasting will likely include clothing type, location, race, age, and social media engagement. We know that such data is readily available from existing social networks like Facebook, Instagram, and TikTok, and this information could be aggregated into personally identifiable data. Perhaps even more significant are the inferences that can be made from the metadata. Cambridge Analytica whistleblower Christopher Wylie, revealed in 2018 how clothing and style preferences can be used to infer the political leanings and ideologies of individual users. For fast fashion companies to implement AI in trend forecasting, the demand for more data to feed algorithm-driven design and marketing efforts will only increase. Without clear AI and data policies, these industry leaders may find themselves battling to win back the trust of their customers if there is a backlash over their AI practices.
Fashion companies like H&M and Zara have already begun to implement trend forecasting to predict market demands, however, there is little transparency about the sources of their data or the details of the metadata. Added to that there are no universal standards for how to implement ethics in the AI development process and internal processes are rarely shared with the public. Multiple polls have shown a rising consumer emphasis on online privacy and increased awareness of the need to balance personalization and privacy. To be sure, consumers enjoy having trendy clothing and benefit from the recommendations of these algorithms, but they are becoming increasingly skeptical as they learn more about how these systems come up with these suggestions. As consumers once campaigned for sustainable supply chain practices, these ethically conscious consumers will undoubtedly advocate for more transparent AI practices. AI-driven fashion companies will face increased scrutiny over their AI practices and have the burden of proof to show that their AI operations do not cause negative social impact. Building ethical AI principles into the AI development cycle can ensure that unintentional outcomes are minimized. Designing ethics into AI operations ensures that the principles of human agency and algorithmic accountability are core features of the business, not an after-thought. Through the operationalization of ethical principles, these tech-savvy fashion titans can realize their AI-driven potential while bolstering trust with an ethically conscious customer base.
Wei-Ann Chang,
Researcher at Aiforgood.asia