Mansha Dhir Deo dives into the use of AI in the fashion industry - from virtual fitting rooms to computer generated models, the fashion industry is shifting. Are we losing the humanity of fashion?
If fashion is a mirror of society, then right now, it’s definitely reflecting pixels. Fashion giant H&M recently announced their plans to create AI-generated model twins—which is one of the many examples of how artificial intelligence is taking over the industry.
From personalised style recommendations to virtual fitting rooms using landmark detection, AI is revolutionising fashion. Deep learning algorithms are analysing vast datasets to predict upcoming trends and optimise production. Designers are using AI to prototype their collections before even a single stitch is sewn. In the near future, hyper-realistic VR fittings may help bring down online return rates, which contribute up to 24 million metric tons of CO? emissions each year, and only 20% of it stemming from damaged goods.
This user-centred approach may help relieve some of the environmental strain caused by the fashion industry. But does it really tackle the root problem? While more accurate trend prediction could reduce unsold inventory, it still feeds into rapid consumption and the cycle of fast fashion, not to mention the environmental impacts of AI itself.
Another common concern with the rise of AI is the looming threat of job automation. As Futurist Sinead Bovell warned, “There are all these nuanced ways that models may get even more exploited.” The rise of AI modelling agencies raises important ethical questions. Who should be the owner of these digital personas? What does fair compensation look like for a digital twin? Should it be standardised, or vary based on each model? Imagine saying yes to one campaign and next thing you know, your face is working overtime for a boutique in San Francisco and you’re not seeing a dime anytime soon.
H&M is not the only brand that ventured into this territory. In March 2023, Levi Strauss & Co. also announced their plans to use AI models to increase diversity, which was widely condemned and labelled as “digital blackface” by critics. In South Korea, Lotte Home Shopping made headlines with their “forever-29-year-old” AI-model Lucy, wherein the company stated that hiring people for ads “costs quite a lot” and “there are changes when (people) age.” It is an intriguing paradox of brands trying to be more inclusive and relatable to consumers by removing real humans out of the frame.
If you’re a fashion designer, release that breath you’ve been holding until now. AI is a tool, not a creator. It is great for brainstorming, producing rendered concepts, maybe even suggesting a good fabric. But the process of designing actual garments only begins there. Designing is about textures, folds, longevity, and harmony. It is about telling stories in the nuances of drapes. Something no algorithm can code.
Besides, AI definitely wouldn’t know that it is impossible to use a half-loop topstitching on low-viscosity rayon (it would snag the fabric).
As AI weaves itself into the fabric of fashion, there is a need to find balance between innovation and authenticity. Responsible use of AI could be a big step towards sustainability and redefining creativity. But if unchecked, it may undermine the human touch that makes fashion an art and raze it to mere soulless efficiency.