Levi’s Use of AI to ‘Increase Diversity’ is Wildly Tone-Deaf
MAR 29, 2023 JARON SCHNEIDER
Last week, Levi Strauss was heavily criticized for its decision to start using digital models generated by artificial intelligence (AI) instead of using that investment to hire more real people, especially since it was done under the guise of “diversity.”
The announcemed partnership with AI-model generation company Lalaland.ai was very likely not received the way the company hoped. Peter Ramsey, best known as the director for 2018’s Spider-Man: Into the Spiderverse, put the general public sentiment the best: “Very efficient, Levi’s! Laziness, cheapness, and cynicism all in one stroke.”
Levi’s would very likely have received some negative pushback if it had simply left the announcement at using AI to generate models, but the emphasis on doing so as a means to increase diversity is particularly tone-deaf and the source of much of the outrage.
When reached for comment, Levi’s told PetaPixel that it hadn’t quite figured out what the integration of Lalaland.ai would look like for the brand yet and also attempted to assuage fears that it would replace humans entirely.
“I’d like to emphasize this will not replace or impact photoshoots, this is purely supplemental to our current ways of working,” a Levi’s representative said.
Unfortunately, that misses the point. Sure, it’s great to hear that humans won’t be entirely replaced by robots, but the fact that Levi’s chose to invest any amount of money in AI-generated, fake people instead of using those resources to find and hire real, actual humans is the problem. It is especially problematic when the stated goal of using Lalaland.ai is to increase the diversity of their models.
“We know our consumers want to shop with models who look like them, and we believe our models should reflect our consumers, which is why we’re continuing to diversify our human models in terms of size and body type, age, and skin color,” the company representative says.
“Lalaland.ai’s technology, and AI more broadly, can potentially assist us by supplementing models and publishing images of our products on a range of body types more quickly, while we coordinate photoshoots with live models and finalize website assets. Additionally, it can help to increase the number of models per product, which is generally one model right now.”
Levi’s says that the goal for this partnership is to allow the brand to expand the number of models that each line of clothing has so that customers have a higher chance of seeing clothes on a “person” that more closely resembles them. As of now, the company basically only has one model per line. Levi’s spokesperson is doing a lot of heavy lifting here to try and soften the blow, but there is a factor that the company isn’t addressing.
PetaPixel asked Levi’s specifically why hiring more models was not a viable solution to this problem. The company confirmed receipt of the question but, unsurprisingly, never provided a response.
Levi’s very likely did not respond to this question because the answer is pretty obvious: there is no good reason why AI would have to be used in this case as opposed to finding real human models. The powers that be at Levi’s determined that it’s more efficient, and likely cheaper, to go this route than to deal with actual people.
The response to what Levi’s is doing has been nearly unanimous: this is not a move to celebrate.
Levi’s seems to at least understand that it didn’t do a good job with communication on this initiative, as seen in a large apologetic addition to its press release dated yesterday. However, the core issues with the company’s plan to use AI remain unanswered.
You know what would be great, Levi’s? How about you hire more photographers of different backgrounds and use them to photograph more models of different sizes and ethnicities? Until you actually act on “commitments to diversity,” you’re just blowing hot air. Using AI to make fake people that look “different” than your typical models isn’t increasing diversity — it’s grandstanding.