• by Sarah Mahoney , Staff Writer @mahoney_sarah, April 9, 2025

H&M is taking a high-profile leap into generative AI, creating digital twins of 30 real-life models for use in ads and social content. The move is already stirring debate.

For some, the move is a savvy blend of tech and trend. For others, it raises thornier questions about authenticity, creativity, and what happens when “real” people are rendered optional.

H&M says the project is deeply collaborative. “We are doing this in close collaborations with other creatives within the industry, as well as model agencies and the models themselves, who will hold the right of usage to their digital twin,” said Jörgen Andersson, H&M’s chief creative officer, in a statement. “We are curious to explore how to showcase our fashion in new creative ways -- and embrace the benefits of new technology -- while staying true to our commitment to personal style. This initiative is about exactly this: exploring the benefits of what generative AI can bring to the creative process… Ultimately, H&M’s purpose is to liberate fashion for the many and to celebrate self-expression in everything we do.”

While other fashion brands -- including Levi’s, Calvin Klein, Prada, and Balmain -- have dabbled in AI-generated models, H&M’s approach takes things a step further. By creating digital twins of real people, the retailer is exploring how these avatars might function not just as visual assets, but as influencers in their own right.

“It’s inevitable this will be adopted quite widely by brands,” says Karen Fielding, chief strategy officer at General Idea, which works with luxury clients including Louis Vuitton and Coach. “How they adopt it, and in which places, will be what’s interesting.”

In practical terms, the appeal is clear: AI speeds up content production, cuts costs, and reduces the need for time-consuming photo shoots. But from a branding perspective, the move is more complex -- especially for a company built on the tension between accessibility and aspiration. “H&M’s business model has always been delivering a Vogue-like kind of brand image at the cheapest possible price,” Fielding says. “To me, the question is, are they still delivering that kind of luxury and image, if they're using this technology rather than real people?”

Ansley Williams, head of influence at Ogilvy North America, sees opportunity in the technology's use, but urges caution. “Does this effort serve a creative purpose? Is this a cost-saving move? Or a different way to appeal to the consumer?” she asks. She notes that Ogilvy has worked extensively with virtual influencers, including Lu from Brazilian retailer Magalu, who has built a large and loyal following by blending digital identity with human-like storytelling.

Williams says consumers deserve transparency about what they’re seeing. “We’re really passionate about disclosure,” she says. But she also believes the real opportunity lies in augmenting -- not replacing -- human connection. “From a creative perspective, people build connection with real people. And when they see real humans being able to interact with clothing and fashion, it adds a human element.”

She predicts a kind of aesthetic whiplash as AI-generated visuals become more common, with a rebound toward messy, imperfect, unmistakably human content. “Maybe we’ll just completely bounce back to what’s true in reality,” she says.

Fielding agrees, especially for luxury. As AI floods the market with infinite amounts of polished but generic imagery, she expects high-end brands to double down on emotional storytelling, craftsmanship, and imperfection. “There will be a move to things that are human -- handmade, imperfect things,” she says. “And those will become more prevalent in luxury.”

But just as most people will never be able to distinguish between a real Hermes bag or a fake, consumers aren’t likely to spend much time poring over the endless stream of fashion images, challenging themselves to spot the imposters. Even if they could, many would still prefer the affordability of mass-produced goods over pricey bespoke options, Fielding adds.

For now, even H&M doesn’t know where this leads. “It’s all about testing and learning,” says Williams. “I know that’s boring. But it’s still so true.”