Every year on the first Monday of May, the most photographed steps in fashion belong to the Met Gala. The looks that climb them take months to build. Human stylists decide what gets worn, how it fits the theme and what story it tells. For everyone else, that role has always gone unfilled. Artificial intelligence (AI) is changing that.
AI Moves Upstream
Predicting what matters in fashion was once the exclusive province of runway editors and professional forecasters. That’s changing. Paris-based Heuritech, which counts Prada, Skims and New Balance among its clients, trains AI on runway data, social media and trend signals to identify what will appear on mainstream racks months before consumers see it, according to NPR.
The company’s algorithms correctly predicted dotted prints, the flat-thong sandal and the color yellow as emerging trends for 2026, all of which showed up on Paris Fashion Week runways.
“We have a massive scale of data that allows our AI to detect early signals of trends, sometimes months before they become visible on the mainstream market,” Noémie Voyer, Heuritech’s fashion expertise lead, told NPR.
Stitch Fix Vice President of Buying Amy Sullivan told NPR that AI recently resolved a merchandising call in hours that previously would have taken weeks. Her team needed to choose between a red and a blue stripe shirt for next spring.
“In the past, you either make a spot decision without really looking at it or you’re requesting samples from vendors overseas that could take weeks and cost a lot of money,” Sullivan said. They put it into AI, got a full on-body image of both options and went with the blue.
Advertisement: Scroll to Continue
Voyer and other forecasters are consistent on one point: the AI handles signal processing, not final judgment. “While our AI is extremely sophisticated, the human aspect still counts,” Voyer said.
From Scroll to Stylist
Stitch Fix generates roughly 43 million outfit combinations daily using its proprietary AI model, according to the company’s newsroom. That shortlist goes to one of its 1,700 human stylists, who make the final selection for each customer’s curated box.
In October, Stitch Fix launched Vision, a generative AI tool that lets clients upload a selfie and a full-length photo and then delivers weekly drops of personalized, shoppable outfit images styled to their body and taste. The images place clients in real-world settings like city streets, holiday parties and tailgate scenes rather than blank studio backgrounds.
“We believe the role of human stylists will stay critical,” Chief Product and Technology Officer Tony Bacos told TechTarget. “Humans aren’t just looking in the rear-view mirror, trying to keep the same thing going.”
The Closet You Already Own
Alta Daily, a fashion app launched in 2025, is solving a different version of the same problem. Most people wear only an estimated 20% of the clothes in their closet, Meta AI reported. Alta Daily’s approach is to digitize the other 80%. Users photograph their entire wardrobe and the app uses natural language prompts to pull from that catalog, recommending outfits for any occasion and showing them on a personal avatar.
The underlying technology is Meta’s Segment Anything Model (SAM), which the Alta team uses to clean and isolate every uploaded garment image. Fashion presents one of the most demanding image datasets. A white sneaker against a white wall or a blue sweater on a wrinkled blue blanket in poor indoor lighting, are conditions that trip up lesser models.
By April, Alta had processed more than 20 million images using SAM without the cost burden of external segmentation APIs, which Alta Founder Jenny Wang described as prohibitive for an early-stage business. The app tracks daily outfits worn. It has a user base across the United States, France, Germany, Mexico and the Netherlands. It starts from what the shopper already has, then layers in occasion context, trend signals and visual confirmation through the avatar.
By 2030, agentic commerce, meaning AI systems that shop on a consumer’s behalf, could account for up to $1 trillion in U.S. B2C retail revenue, according to McKinsey.
The styling layer sits at the front of that stack. It decides what a shopper sees, considers and ultimately buys. At the Met Gala, that decision belongs to a person with a point of view and a relationship with the client. In the rest of retail, AI is learning to do the same job.
For all PYMNTS AI and digital transformation coverage, subscribe to the daily AI and Digital Transformation Newsletters.
