Fashion’s AI future takes shape at China’s biggest AI gathering

The World Artificial Intelligence Conference (WAIC) has, in just a handful of years, become China’s answer to South by Southwest (SXSW), except bigger, faster-growing, and considerably more state-backed. This year’s edition ran from July 17 to 20 in Shanghai, drawing roughly 300,000 attendees and over 1,100 exhibitors. For four days, the entire city descended into a kind of frenzy. Bars, hotels, and subways were plastered with AI banners, while FOMO reached fever pitch. Embodied AI alone took up an entire floor this year, making the case that robots are good for more than dancing; they can pour tea, stir-fry, and even give shoulder massages. Meanwhile, the fashion-adjacent AI story was mostly an internal one: agents streamlining CRM systems, supply chain dashboards learning to forecast demand, and workflows quietly becoming automated behind the scenes. It was AI doing a company’s paperwork — useful, but invisible to anyone outside the organization. This year, AI also moved beyond the back office and into the material world — onto hair strands, fabric samples and factory floors. In doing so, it exposed a tension that CRM automation never had to confront: where, exactly, should the human stand alongside the machine? AI steps out from behind the scenes The clearest marker of AI moving from behind-the-scenes automation to creative and product innovation was L’Oréal’s appearance at WAIC, making it the first global beauty conglomerate to participate in the conference’s history. Weeks earlier, at VivaTech in Paris, L’Oréal had unveiled an AI-powered Hair Digital Twin as part of its 10th anniversary “Beauty Odyssey” showcase. This 3D model of hair is built on what the company calls the world’s largest hair science database, capable of simulating a formula’s effect on any hair texture at 100x a physical lab’s resolution and 3x its speed. In Shanghai, the same technology reappeared almost immediately, rebranded “Tech is in the Hair,” sitting a few booths down from robotics startups. L’Oréal’s other booth featured new in-store beauty associates practicing their sales skills on an AI-simulated customer. The coaching tool is designed to accelerate training and give managers clearer visibility into staff performance. The human stays; the ramp-up time shrinks. When automation doesn’t add up In manufacturing, the AI dynamic is closer to direct substitution. A source who works closely with garment manufacturers told Jing Daily that nearly every factory owner he knows is shopping for robots right now, simply because fewer people want factory work. Even Vietnam, which acted as the fallback when Chinese labor got expensive, is starting to see a shortage. However, the return on investment isn’t there yet. A sewing robot can run 500,000 RMB to 700,000 RMB ($73,820 to $103,350) — for that price, human labor is still the cheaper option, especially in much of Asia. Another source, who’s spent years around sewing automation, made the case that the problem is more solvable than people think. Unlike couture, ordinary garments involve a fairly standardizable set of variables — fabric weight, texture, and thread type — making them easier to model and automate. Alibaba tested this idea in 2020 with a company called Rhino Smart Manufacturing, built around AI-driven textile manufacturing. However, scaling it proved difficult. The bottleneck, in other words, was not the technology itself but the cheap labor that’s still hard to beat. One booth suggested where AI-driven manufacturing might be heading. Style3D (凌迪科技) has spent a decade building simulation software to model how fabric drapes and folds, a notoriously difficult problem given how unpredictably soft materials behave. At WAIC, the company showed a new platform called SynReal World, repurposing that same engine to generate synthetic training environments where robots can learn physical tasks without costly real-world trial and error. By tackling garment simulation first, everything downstream was comparatively easier. Drawing the line between human and machine When it comes to services, the industry is still deciding how close AI should get to the customer without replacing the person meant to greet them. In manufacturing, the question is more straightforward: whether automation makes financial sense. Somewhere between a digital hair strand and a sewing robot still too expensive to justify, this year’s WAIC made one thing clear: fashion’s AI story is no longer just about efficiency. It is about where, exactly, companies are willing to let machines stand in for people — and where they aren’t.

Fashion’s AI future takes shape at China’s biggest AI gathering

How Bottega Veneta weaved China’s forever it bag

Upgrade to Pro

Sectors

Insights

Company

Fashion’s AI future takes shape at China’s biggest AI gathering

How Bottega Veneta weaved China’s forever it bag