AI virtual try-on: from static catalogs to conversational discovery
AI virtual try-on in beauty shopping refers to interactive, machine-learning powered experiences that let consumers see cosmetics or fragrances on themselves digitally, while receiving tailored recommendations through conversational interfaces that connect product data, visual simulation and personal preferences into a single, guided discovery journey. This is no longer a novelty; it is becoming the new front door to beauty. Brands see that AI is permeating innovation across industries, and beauty players are rushing to keep pace by integrating AI across personalization and production. The most interesting shift is not that AI exists, but that it is moving straight into the shopping flow, where consumers expect advice, not ads. Virtual try-ons are the clearest signal that beauty is turning into a dialogue between the shopper, the product, and the algorithm.
L’Oréal’s CreAItech and OpenAI: marketing system meets AI product discovery
L’Oréal’s partnership with OpenAI is a bet that beauty personalization AI must be embedded in the entire value chain, not bolted onto the front end. The company is expanding its CreAItech generative marketing system with access to OpenAI’s latest models, using them to produce images and video for social, e-commerce and paid channels at scale. This is more than a creative upgrade. L’Oréal plans to feed up-to-date product information directly into the models behind ChatGPT so AI answers about its products pull from brand-controlled notes rather than scattered web data. That move is unapologetically about AI product discovery: whoever shapes what consumers see in AI-led search will shape which brands they consider. According to Gartner’s Greg Carlucci, nearly one in five consumers now use generative AI tools to find information, and personal care brands are seeing website traffic drop as discovery shifts into these tools.
Maybelline’s ChatGPT try-on feature: turning advice into action
If CreAItech is L’Oréal’s engine, Maybelline’s AI virtual try-on inside ChatGPT is the showroom. The brand will offer Makeup Virtual Try-On via ChatGPT, powered by L’Oréal’s ModiFace technology, with the explicit goal of increasing product discovery through AI-powered conversations. Plans call for the virtual try-on experience to sit directly inside ChatGPT, so a user asking how to find a red lipstick can move from guidance to simulated application without leaving the chat interface. This is where beauty personalization AI earns its keep: an assistant that knows the catalog, understands the user’s request, and can visually apply products removes the friction between interest and trial. L’Oréal is already testing paid advertising in ChatGPT with CeraVe, SkinCeuticals, and Garnier in the US, though it stresses that results are still early. The commercial logic is blunt: the more discovery happens here, the more brands must fight to be present in this stream.
Beyond one brand: AI personalization and production across beauty
The beauty industry’s AI push is wider than one conglomerate, and that matters because it shows a structural change rather than a marketing trend. Beauty brands are using AI to personalize consumer experiences, such as an AI-powered fragrance discovery model that turns users’ aesthetic preferences on Pinterest into Jo Malone London scent suggestions. On the skin side, Amorepacific’s AI-powered Facial Aging Map visualizes wrinkle and hyperpigmentation changes and generates regional aging skin patterns that can inform personalized care by geography. At the production level, Unilever’s work with Accenture on AI-powered digital twins has already cut waste by 20% and boosted capacity by 10% at one factory, with over 40 new digital twins planned over 18 months. These moves share a worldview: AI should not sit in a single app. It should connect product development, operations and front-end personalization so that every new SKU and every new ad is born AI-aware.
The next phase: AI-led beauty discovery as the default
L’Oréal’s partnership with OpenAI reaches into research as well, with access to GPT-Rosalind to map the skin microbiome as part of its focus on La Roche-Posay. That detail matters because it hints at a future where AI shapes not only how products are sold but what products exist. If AI product discovery increasingly runs through conversational systems, the brands that train those systems with detailed product data and real science will own the narrative. Lancôme and Kérastase are already working to amplify discovery within ChatGPT using enhanced signals, and L’Oréal has trained 70,000 staff in AI use, treating it as a company-wide capability build rather than a niche experiment. The takeaway is stark: beauty shoppers will not go back to static lists once they taste guided, AI virtual try-on experiences. Brands that hesitate will find their products absent from the very conversations that now start the buying journey.
