H&M Group

H&M uses machine learning models to scan social media platforms, fashion blogs, search trends, and internal sales data to identify emerging fashion trends in real time, feeding those insights into product design and buying decisions.

H&M's trend analysis AI uses natural language processing and machine learning to process large volumes of unstructured data from sources including Instagram, TikTok, fashion blogs, Google Trends, and point-of-sale data. The system identifies emerging colors, styles, and materials and surfaces these patterns to design and buying teams. H&M employs more than 200 data scientists, analysts, and engineers to develop and run these analytics capabilities.

H&M deployed AI-generated digital twins of 30 real human models, in partnership with AI provider Uncut, to produce product imagery and marketing content for social media and advertising campaigns. The first set of images was published in July 2025, with all AI-generated content watermarked and labeled for transparency.

H&M announced the initiative in March 2025, partnering with Swedish tech firm Uncut to photograph models extensively from multiple angles, movements, and lighting conditions, then process that data through generative AI to produce lifelike digital replicas. The resulting images are used in place of traditional photoshoots and are clearly watermarked as synthetic. H&M states that models retain ownership of their digital twins and that no likeness is used without explicit consent.

H&M uses AI algorithms to recommend products to shoppers based on their purchase history, browsing behavior, and demographic data, both on the H&M website and app and in physical stores using RFID technology. H&M hosted a public Kaggle competition in 2022 to improve its recommendation engine.

H&M's recommendation system analyzes customer purchase histories, browsing patterns, and regional sales data to surface personalized product suggestions in digital channels. The company extended these online recommendations to physical stores using RFID technology, allowing customers to receive in-store merchandise suggestions selected by algorithms. In 2022, H&M Group partnered with Kaggle to run an open data science competition using real H&M transaction data to develop better recommendation models, with the goal of improving its production recommendation engine.