# ChatGPT Launches Virtual Try-On Feature for Festive Shopping OpenAI integrates generative AI to visualize clothing on users for improved shopping experiences. - **Source:** [Sci-Tech News | Latest Technological Developments | The Hindu](https://www.thehindu.com/sci-tech/) - **Published:** 2026-10-28 - **Topic:** [Marketing](https://brifing.si/c/marketing) - **Original article:** - **On Brifing:** https://brifing.si/s/99a0682c-c8bc-4b75-92b8-830842dfe5e4 ## Generative AI Replaces Mannequins Retail has spent over 15 years attempting to solve the online fit problem through robotic mannequins and 3D body modeling. Companies like Fits.me and Metail struggled with high costs and scalability. OpenAI now shifts this approach by using generative AI to place clothing directly onto user-provided photographs, bypassing complex body modeling. ## Technical Implementation and Launch Launched globally on October 1, 2026, the Virtual Try On feature utilizes the ChatGPT Images 2.5 model. Users can trigger the feature via a button on supported product listings or by uploading external screenshots. The system generates a visual representation of the garment on the user, aiming to assist with initial purchase decisions. ## Managing User Expectations OpenAI product manager Aarush Selvan emphasizes that these generated images serve as visual references rather than fit guarantees. The technology does not reproduce the person or garment with absolute precision. Users are advised to consult merchant-provided size charts, product details, and return policies before finalizing any purchase on external retail websites. ## Inclusivity and Bias Mitigation To address concerns regarding body representation, OpenAI conducts evaluations across diverse races, skin tones, genders, and religions. The system applies existing safety guardrails to ensure the generated output remains faithful to the user's features. However, the company maintains that the output should not be treated as a perfect, verified reproduction of the individual. ## Handling Missing Visual Data When users upload screenshots that lack specific details, the generative model attempts to fill in missing information based on its training data. OpenAI warns that these AI-generated additions are not verified facts about the garment. Shoppers should rely on official product descriptions for accurate information regarding hidden seams or specific fabric textures. ## Limitations in Ensemble Styling The current feature supports single-item try-ons rather than full outfits. While users can upload multiple images to request a composite view, there is no dedicated multi-product try-on tool available. This requires manual workarounds for complex ensembles, such as matching a Kanchipuram saree with specific temple jewelry or coordinating a groom's veshti. ## Data Privacy and Personalization Reference photos used for try-ons are retained to facilitate future sessions, though users can delete or replace them via settings. Generated images are saved to the Library and remain private by default. Product information for saved items is refreshed dynamically when clicked, ensuring that pricing and stock status remain as current as possible. ## Future of Festive Retail As the wedding season approaches, this technology aims to move the trial room experience to the consumer's phone. By integrating discovery, try-on, and saving features, OpenAI positions ChatGPT as a starting point for shopping journeys. The ultimate challenge remains whether these digital visualizations can satisfy the social approval required for traditional events. ## Sources covering this story - https://www.thehindu.com/life-and-style/fashion/chatgpts-virtual-try-on-moves-the-trial-room-to-your-phone-this-festive-season/article71610774.ece — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._