AI Could Have Terrifying Impacts on Fashion Models, New Research Shows

Artificial intelligence, often abbreviated as AI, has been a topic of conversation in nearly every industry worldwide, and fashion is no exception. While AI's impacts on a large scale are still unknown, there is one place in fashion where we are already seeing change — and, unfortunately, subsequent problems. In a new report released by Cornell University's ILR School and Worker Institute, authors Alexandra Mateescu, Zoë West, and Sanjay Pinto found that AI is already changing the fashion modeling industry, and the scope of this shift extends far beyond job security.

According to the research, “Frankensteining,” or altering a model's image, has long been standard practice in fashion. Generative AI is not only making things easier but also more profound. “A single photoshoot or booking with a model can be used by client companies to capture images, body scans, and/or measurements that can then be easily manipulated with AI technologies, allowing companies to extract additional profit from this data without needing to re-hire or further compensate the model,” the report claims. Moreover, models are finding that agencies aren't equipped for the rapid increase in usage, and therefore, are put in situations where their likeness is manipulated or used without their knowledge.

Of course, it's not all murky. Some brands, such as H&M, have been very clear about their intentions to utilize AI for fashion modeling, but that doesn't alter the impact of the practice. Many fashion models are already subject to a power dynamic in which the brands and the agencies have significant sway over their livelihoods. Specifically, the report finds that e-commerce and fit models (models that brands work with to test samples) are subject to this type of AI usage, where brands have the potential to create the illusion of youth, racial, and body equity through AI models, the most.

Last week, ultra-fast fashion brand Shein came under fire after an image of Luigi Mangione, the man accused in the shooting of UnitedHealth CEO Brian Thompson, who has since pleaded not guilty to all federal charges, appeared on its site. According to the BBC, “The image in question was provided by a third-party vendor and was removed immediately upon discovery.” Shein told the outlet that they are "conducting a thorough investigation, strengthening our monitoring processes, and will take appropriate action against the vendor in line with our policies."

The image showed what appeared to be a smiling Mangione wearing a men's floral button-up shirt. While Shein has not confirmed that the image was, in fact, AI-generated, it seems clear that Mangione did not recently model for the brand, as he has been in jail awaiting trial since December 2024.

Whether the product listing was a mistake by a third party or intentional, it highlights a serious lack of discernment in this practice. Outside of the obvious trivialization of a serious court case (and cultural conversation about healthcare in the United States), the way Mangione's likeness got onto that page was not through a paid partnership. It means that models, influencers, or really, any of us could potentially find our likeness being used to sell something we don't even know about, or possibly believe in.