top of page
  • Writer's pictureSarah Ruivivar

AI Alters Model's Ethnicity: Progress or Setback?

Image credits: TikTok user @shereenwu

Picture this: you're a model, strutting down the runway, basking in the camera flashes.

The next day, you find your photo on social media, but there's a catch - your face has been edited to appear as a different ethnicity. Sounds like a nightmare, right? Well, this is precisely what happened to Shereen Wu, a Taiwanese American model.

In a TikTok video viewed 1.8 million times, Wu revealed that a photo of her from a recent fashion show was digitally altered by the designer, Michael Costello, making her appear white. Costello, however, denied the allegations, stating the image was "fan art" and he was unaware of the alteration.

This incident raises questions about the ethical use of AI in the fashion industry. A decade ago, models were protesting against their bodies being photoshopped to unrealistic proportions. Today, the industry grapples with the ethical implications of AI, a technology that can perpetuate racist and sexist stereotypes.

Levi's recently announced its use of AI-generated models on its website to promote diversity. However, critics argue that hiring human models of diverse ethnicities would be a more authentic approach.

Susan Scafidi, academic director of Fordham’s Fashion Law Institute, warns that AI could potentially erase a model's race, "turning back the clock on the fashion industry’s progress toward diversity on the runway."

The incident involving Wu highlights the urgent need for regulations protecting models' rights. Currently, copyright law protects photographers, but not the models they shoot.

As we navigate the future of AI in the fashion world, it's crucial to ensure that technology promotes diversity and respect, rather than erasing individual identities. After all, in the words of Wu, "The very thing that makes us human is our ability to create something beautiful." Let's not allow AI to twist this beauty into something potentially ugly.

Made with TRUST_AI - see the Charter:

11 views0 comments


bottom of page