H&M's digital twin models: consent, payment and disclosure examined

H&M created digital twins of 30 real models in 2025, paying each model an upfront fee for the rights to their digital likeness and disclosing AI-generated imagery through visible watermarks. The Swedish fast-fashion retailer's approach offers one of the first large-scale commercial examples of how fashion businesses are navigating consent, compensation and transparency as generative AI transforms product photography and marketing.

How did H&M create and deploy digital model twins?

H&M worked with an external technology provider to scan and create digital replicas of 30 professional models during a controlled studio session in early 2025. Each model signed a written consent agreement that specified how their digital likeness would be used, the duration of rights granted, and the territories where the images could appear. The digital twins were then deployed across H&M's website and mobile app to model garments in product photography, replacing some traditional photoshoots for catalogue imagery.

The retailer applied the digital twins primarily to basic product categories such as T-shirts, jeans and knitwear where standardised poses and neutral backgrounds are the norm. More complex editorial campaigns and seasonal lookbooks continued to use traditional photography with live models. This tiered approach allowed H&M to test the technology on high-volume, repeatable imagery whilst preserving the creative flexibility required for brand storytelling.

What payment model did H&M use for digital likenesses?

H&M paid each of the 30 models a one-off fee for the rights to use their digital likeness over a defined period, rather than per-image royalties or residual payments. The exact fee amounts were not publicly disclosed, and the contracts varied based on the model's experience and the scope of rights granted. Industry observers noted that this upfront payment structure mirrors traditional buyout agreements common in advertising, where models receive a lump sum in exchange for extended usage rights.

The absence of residual payments has drawn attention from model advocacy groups, who argue that digital twins can generate unlimited images over time, potentially diminishing future earning opportunities for the individuals involved. Unlike a traditional photoshoot where a model's participation is finite, a digital twin can be reused indefinitely within the contract terms, creating an imbalance between the one-time payment and the ongoing value extracted from the likeness.

How does H&M disclose AI-generated model imagery?

H&M added visible watermarks and text labels to product pages featuring AI-generated model imagery, indicating that the images were created using digital twin technology. The disclosures appear as a small icon or text notice near the product images, informing customers that the model is a digital replica. This transparency measure aligns with emerging regulatory expectations in the European Union and voluntary industry guidelines that call for clear labelling of synthetic media.

The watermarks serve a dual purpose: they inform consumers that the imagery is not a traditional photograph, and they help build trust by demonstrating that the retailer is not attempting to pass off AI-generated content as real. However, the size and prominence of the disclosures have varied, and some consumer advocates have questioned whether the labels are sufficiently noticeable to the average shopper browsing quickly through product listings.

What are the business and efficiency benefits for H&M?

The digital twin programme offers H&M several operational advantages. By reducing the need for physical photoshoots, the retailer can lower costs associated with studio rentals, photographer fees, styling, hair and makeup, and model day rates. Digital twins also enable faster turnaround times: new garments can be photographed on virtual models within hours rather than the days or weeks required to schedule and execute a traditional shoot.

The technology also provides greater flexibility for global markets. H&M can adapt digital twin imagery to different regional preferences by adjusting backgrounds, poses or styling without re-shooting, and can scale production to match the velocity of fast-fashion product cycles. For a retailer that launches thousands of new items each season, the ability to generate consistent, on-brand imagery at speed represents a meaningful competitive advantage.

• Reduced photoshoot costs: no studio, crew or travel expenses for repeat product shots
• Faster content production: digital twins can model new items within hours
• Scalability: one digital twin can appear in unlimited images across markets
• Consistency: standardised poses and lighting for catalogue imagery

What legal and ethical questions remain unresolved?

H&M's digital twin initiative highlights several unresolved legal and ethical questions in the fashion industry's adoption of AI-generated imagery. The one-off payment model raises concerns about fair compensation: if a digital twin generates thousands of images over multiple years, does a single upfront fee adequately reflect the value of the model's likeness? There is currently no industry standard or regulatory framework governing how digital likeness rights should be priced or structured over time.

Another open question concerns the long-term control of digital likenesses. If a model's contract with H&M expires, what happens to the existing images that were created using their digital twin? Can those images continue to circulate, or must they be removed from the retailer's archive? These questions of digital rights management are still being negotiated between models, agencies, retailers and legal advisors, with few precedents to guide the parties.

Transparency and consumer perception also remain areas of active debate. Whilst H&M has implemented watermarks, the effectiveness of these disclosures in shaping customer understanding is unclear. Some research suggests that consumers may not notice or understand synthetic media labels, whilst others argue that clear disclosure builds trust. As AI-generated imagery becomes more widespread, the fashion industry will need to establish clearer norms for how and when to inform customers about the nature of the content they are viewing.

What does H&M's approach signal for the wider industry?

H&M's digital twin programme represents an early commercial template for how large fashion retailers might integrate AI-generated model imagery into their operations whilst addressing consent, payment and transparency. By securing written agreements, paying upfront fees and labelling synthetic media, H&M has taken steps that many industry observers view as more responsible than simply deploying AI without disclosure or compensation.

However, the programme also underscores the need for evolving industry standards. As more retailers experiment with digital twins and synthetic models, questions about fair compensation, long-term rights and consumer understanding will only intensify. The fashion industry, model agencies, trade unions and regulators will need to collaborate to establish frameworks that protect individual rights, ensure transparency and allow businesses to realise the efficiency gains that AI offers. H&M's 2025 initiative provides a real-world case study from which the broader industry can learn, but it is unlikely to be the final word on how digital likenesses are managed in fashion.