H&M
AI digital twins of models for creative production
In 2025, H&M created with the Swedish studio Uncut AI digital twins of around thirty real models, who keep ownership and a revenue share, to produce ad, social, and e-commerce visuals without reshoots; first campaign live on July 2, 2025.
Objective
Produce fashion visuals (ads, social, e-commerce) without multiplying physical shoots, while keeping real models, their rights, and their pay.
The deployment
H&M digitized around thirty real models to turn them into AI-generated digital twins, with the Swedish tech studio Uncut. Each model is photographed from several angles and lighting setups to capture their features and movements, then Uncut produces a usable replica. The twin is then used to compose product visuals and campaign images without organizing a new physical shoot: if a model is already working elsewhere, her twin can wear another collection the same day. The framework is designed to defuse the rights issue: the model keeps ownership of their twin, can rent it to other brands, including competitors, and is paid on a basis aligned to the standard use of their image. AI-produced content is labeled on social media. The first public campaign went live on July 2, 2025 around denim, with twins staged against the backdrop of fashion capitals, co-created with a dedicated team (photographer Johnny Kangasniemi, model Vanessa Moody) and accompanied by a making-of film. H&M announced further AI drops for fall 2025.
Results Proof C
Official H&M Group press release (primary source) confirming the launch, corroborated by several established news outlets naming the brand and the partner Uncut. No financial results isolating the impact of the setup, so no level A; no quantified platform case study, so no level B.
How it works
Inferred typical approachThe internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.
How it runs, concretely
For ops teams-
1Digitize the model H&M creative team and AI partner
Photograph each model from several angles and lighting setups to capture their features and movements.
-
2Build the digital twin AI partner (Uncut)
Uncut generates the usable AI replica from the captured data.
-
3Frame rights and pay H&M and modeling agencies
Contract the model's ownership of the twin and pay aligned to the standard use of their image.
-
4Generate the visuals H&M content team
Compose product and campaign images by dressing the twin, without a physical reshoot.
-
5Publish with labeling marketing / social
Publish on social and e-commerce, clearly marking the content as AI-generated.
The fidelity of each model's initial scan (angles, lighting, features, movements) and the validity of the rights framework. If the scan is poor, the twin is not usable; if the rights are not framed, the visual is not publishable.
How your customers perceive this type of use
Sourced studiesUn ecart net separe les annonceurs des consommateurs : 77% des annonceurs voient l'IA positivement contre 38% des consommateurs (Yahoo/Publicis, 2024). Les mesures implicites confirment le rejet declare : en EEG, les pubs generees par IA produisent une activation memorielle plus faible que les pubs traditionnelles et sont decrites comme agacantes, ennuyeuses et confuses (NIQ, 2024). La disclosure a un effet ambivalent : elle augmente fortement la confiance quand elle est remarquee (Yahoo/Publicis), mais 27% des jeunes consommateurs disent faire moins confiance a une entreprise dont la pub est creee par IA (IAB, 2024).
Acceptance conditions
- Une disclosure visible : quand la mention IA est remarquee, la confiance globale envers l'entreprise augmente de 96% (Yahoo/Publicis 2024)
- Une qualite visuelle suffisante : les visuels IA de basse qualite augmentent l'effort cognitif et distraient du message (NIQ 2024)
Red lines
- Le contenu IA non declare puis identifie : 72% des consommateurs disent que l'IA rend l'authenticite difficile a etablir (Yahoo/Publicis 2024) et les marques utilisant des pubs IA sont plus souvent jugees inauthentiques ou non ethiques par les consommateurs que par les dirigeants (IAB 2024)
- Les mannequins et personnes generes par IA : 46% des consommateurs n'en veulent pas dans la publicite, l'inquietude premiere etant les standards de beaute irrealistes (Attest 2025)
Sources: Yahoo / Publicis Media (terrain Ebco) 2024 · IAB (avec Attest) 2024 · NIQ (NielsenIQ) 2024 · Attest 2025
How to replicate
Inference, not sourcedData prerequisites
- Agreement of real models and digitization data (multi-angle scans, lighting, movements)
- Asset library to store and version the twins
- Product catalog and style reference set to dress on the twins
Org prerequisites
- Legal framework for image-rights transfer and model pay
- AI Act compliance: systematic labeling of AI-generated content
- Alignment between creative direction, legal, and the production studio
Possible stack
- Specialized digital-twin partner (e.g. Uncut)
- Generative image models
- Digital asset management (DAM) solution
The plan, step by step
- Step 1Select a small group of volunteer models and frame their rights and payDeliverable: Signed transfer and pay contracts
- Step 2Digitize the models from several angles and lighting setupsDeliverable: Reliable source data per model
- Step 3Produce the digital twins with an AI partnerDeliverable: Library of usable twins
- Step 4Generate a first test campaign on one product categoryDeliverable: Campaign visuals produced without a reshoot
- Step 5Label and publish on social and e-commerce, then measure perceptionDeliverable: Compliant published content and audience feedback
First step: Select a few volunteer models and write the rights and pay framework before any digitization: this is what makes the setup publishable and defensible.
Sources
- S1 H&M continues its exploration of creativity with AI Primary archive pending
- S2 H&M Releases First Images With AI Digital Twins Established press archive pending
- S3 H&M turns to AI 'digital twins' in new campaign, as fashion grapples with blurred realities Established press archive pending
- S4 H&M drops first AI 'digital twins' to showcase denim collection Secondary archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.