How is AI used in physical fashion stores?
From associate assistants and clienteling to computer vision for stock and checkout, AI is moving onto the shop floor. What works, what it requires and where privacy law draws hard lines.
KEY TAKEAWAYS Summary by the editors
- AI in physical fashion stores is used mainly for associate assistance, clienteling, inventory and replenishment, computer vision for shelf and stock monitoring, and loss prevention.
- Levi Strauss built a generative AI assistant for store staff, called Stitch, with Google Cloud; Fortune reported in March 2026 that it was live in more than 70 US stores after a ten-store pilot.
- Since 2 February 2025, the EU AI Act has prohibited AI systems that infer emotions of people in the workplace, except for medical or safety reasons, which rules out emotion monitoring of store staff.
- The UK ICO states that organisations must identify both a lawful basis and a separate condition for processing special category biometric data in biometric recognition systems.
- Store AI depends on accurate product, stock and customer data, and on associates actually trusting and using the tools.
AI in physical fashion stores helps associates answer questions and serve customers, keeps stock and replenishment accurate, and supports checkout and loss prevention through computer vision. The most practical current uses support staff rather than replace them, while camera-based uses involving people face strict limits under privacy and AI law.
What are the main uses of AI in fashion stores?
| Use case | What it does | Data required | Main constraint |
|---|---|---|---|
| Associate assistant | Answers product, procedure and policy questions in natural language | Product information, procedures, training content | Accuracy of underlying content |
| Clienteling | Gives staff customer history and suggested products for personal service | Unified customer profile with consent | Data protection and staff adoption |
| Inventory and replenishment | Predicts store-level demand and suggests transfers or reorders | Sales, stock by size, deliveries | Data accuracy at store level |
| Shelf and stock vision | Detects gaps, misplaced items or planogram issues from images | Camera images of fixtures, product reference images | Installation cost, coverage |
| Footfall analytics | Counts visitors and analyses movement without identifying them | Sensor or camera data, ideally anonymised | Privacy by design |
| Loss prevention | Flags unusual events at checkout or exits | Video, transaction data | Legal limits on biometrics, false positives |
How does AI help store associates?
Store staff are expected to know a large, fast-changing assortment, store procedures and return rules, often with limited training time. Generative AI assistants let them ask questions in plain language and receive answers drawn from approved internal content.
Levi Strauss is a well-documented example. According to Fortune, its assistant, Stitch, began as an employee idea at an internal hackathon and was built with Google Cloud using Gemini models. Associates use it on tablets or smartphones to ask, for example, how two jean fits differ, how to process a return or how a procedure works. After a pilot in ten stores in late 2025, Fortune reported in March 2026 that it was available in more than 70 US Levi's stores, with more locations and languages planned. Jason Gowans, the company's chief digital and technology officer, told Fortune that stores with access showed an eight-point improvement in consumer satisfaction compared with those without, which he described as giving the company "some intuition that there's real value here".
That framing is appropriately cautious: a store comparison is suggestive, not proof, and results depend on how well the underlying product and procedure content is maintained.
What is AI-supported clienteling?
Clienteling means personal, relationship-based selling, traditionally strongest in luxury and premium retail. AI tools support it by giving associates a view of a customer's purchase history, preferences and sizes, suggesting products that complement past purchases, and drafting follow-up messages. The value depends on a unified customer profile that links online and in-store activity and on recorded consent for using that data. Without both, clienteling tools show incomplete pictures and create compliance risk.
How does AI improve store inventory and replenishment?
Stores lose sales when the right size is not on the floor, even if it is in the stockroom or a nearby store. Machine learning models can forecast demand at store and size level, recommend replenishment from the stockroom, and suggest transfers between stores before items sell out or end up marked down. These models depend on accurate stock records; where the system believes items are present that are actually missing, forecasts and recommendations degrade quickly. That is why many retailers treat item-level stock accuracy, for example through regular counts or tagging, as a precondition for store AI rather than an optional extra.
How is computer vision used in fashion stores?
Computer vision analyses images from cameras. In stores it can detect empty or untidy fixtures, check visual merchandising against plans, support self-checkout and count footfall. Uses that analyse objects and fixtures are legally simpler than those that analyse people.
Where systems analyse people, the law sets firm boundaries. In the EU, Article 5 of the AI Act has, since 2 February 2025, prohibited AI systems that infer the emotions of people in the workplace (except for medical or safety reasons), biometric categorisation systems that infer sensitive characteristics such as race, political opinions or sexual orientation, and the creation or expansion of facial recognition databases through untargeted scraping of facial images from the internet or CCTV footage.
Facial recognition of shoppers, for example to identify suspected shoplifters, also raises data protection issues. The UK Information Commissioner's Office states in its biometric guidance that organisations must identify a lawful basis and a separate condition for processing special category biometric data, and that if no valid condition can be identified, the data must not be processed.
What do fashion stores need to make AI work?
- Accurate store-level data: stock by size and location, which often requires item-level tracking and disciplined processes.
- Maintained content: product information, procedures and policies that the assistant can rely on.
- Devices and connectivity: associates need reliable mobile devices and in-store network coverage.
- Associate involvement: tools designed with staff, as in the Levi's hackathon origin, are more likely to be used.
- Governance: data protection impact assessments for any customer data or camera use, and clear rules on staff monitoring.
What are the limits of AI in physical retail?
Store AI is constrained by hardware cost, the difficulty of keeping store-level data accurate and the variety of store formats in a single estate. Associate tools only help if answers are correct; a confident wrong answer about a returns rule damages trust with staff and customers alike. McKinsey's State of Fashion 2026 identifies a "workforce rewired" theme, with technology reshaping positions and creating new roles. For stores, the realistic near-term picture is AI as a support for informed, confident associates rather than a substitute for them.
Frequently asked questions
How do fashion retailers use AI in stores?
Mainly for associate assistants that answer product and procedure questions, clienteling, inventory and replenishment, computer vision for shelves and checkout, and loss prevention. Most proven uses support staff rather than replace them.
What is an AI assistant for store associates?
It is a generative AI tool that lets staff ask questions in plain language and receive answers from approved product, policy and training content. Levi Strauss reported deploying such a tool, Stitch, in more than 70 US stores by March 2026.
Is facial recognition allowed in fashion stores?
It is heavily restricted. In the UK, the ICO requires a lawful basis and a separate condition for processing special category biometric data. In the EU, the AI Act prohibits certain biometric uses, such as untargeted scraping to build facial recognition databases. Legal advice is essential before any deployment.
Can retailers use AI to monitor employees' emotions?
Not in the EU. Since 2 February 2025, Article 5 of the AI Act prohibits AI systems that infer emotions of people in the workplace, except for medical or safety reasons.
One edition every weekday morning. Read in five minutes. Free for industry professionals.
