Model Licenses Her Likeness to AI Firm as Fashion Industry Adapts to Generative Tools

After years of freelance work in magazines and campaigns, fashion model Hannah James has licensed her name, image and likeness to an artificial intelligence startup as a hedge against shrinking commissions and a shifting industry.

James told the Daily Mail that earnings fell during the COVID-19 pandemic and dropped further with the arrival of generative AI, prompting her to seek new ways to monetize her work. She now works with Kartel.ai, a company that builds AI models trained on a model's photographs to produce advertising images and short video content that feature a hyper-realistic digital representation of the talent.

Kartel's co-founder, Ben Kusin, described the company's approach as collaborative. Rather than replacing human creative direction, Kartel says it licenses models, then works with creative directors, producers and the models themselves to generate branded assets without needing the model to be physically present on set. The firm builds LoRA (Low-Rank Adaptation) models — a technique that adapts foundation models to recognize specific people or objects — and uses that training to produce stills and short-form video for campaigns.

Kusin said the process reduces the time and cost associated with traditional shoots and offers greater scheduling flexibility for brands and talent. He acknowledged that adoption of AI tools can lead to job losses in some areas, but framed Kartel's model as a way to reskill and provide revenue streams for creatives, arguing that humans still control scripting, storyboarding and approvals.

James said she welcomed the opportunity to profit from her image without the physical and mental demands of frequent travel and shoots. "I think that this is a really cool opportunity to be able to monetize what we do and still be able to make money, make it fast, and actually get paid whenever we're supposed to get paid," she told the Daily Mail.

The move by James reflects a broader industry shift. Companies such as FanPro, Lalaland.ai, Deep Agency and The Diigitals have built businesses around AI-generated models, and established brands have begun experimenting with the technology. A campaign by fashion label Guess that used AI-generated models ran in the August issue of Vogue and drew immediate criticism for not employing real models and for perpetuating narrow beauty standards.

Experts interviewed about the trend say the technology is already influencing how fashion imagery is produced and could further reshape employment in creative industries. Nell Watson, a fellow at Singularity University and author of Taming the Machine: Ethically Harness the Power of AI, told the Daily Mail she expects traditional modeling to decline in many segments, with luxury brands continuing to use live talent while fast-fashion and smaller retailers increasingly opt for synthetic imagery to lower costs and accelerate content production.

A Stanford study released in August found that adoption of generative AI corresponded with a 13 percent decline in employment for early-career workers in industries where the tools were used, underscoring potential labor-market effects beyond modeling.

Legal protections for likeness and image rights remain uneven, attorneys say. Yelena Ambartsumian, founder of AMBART LAW, which focuses on privacy, AI governance and intellectual property, told the Daily Mail that a model's ability to control or profit from an AI-generated likeness depends on recognition and bargaining power. High-profile figures can pursue claims under federal trademark law, including the Lanham Act, if a generated image falsely implies endorsement. Lesser-known models often must rely on a patchwork of state privacy and publicity statutes or contract negotiations to secure protections.

Ambartsumian and others warn of a "wild west" period while legal standards catch up to technological advances. Determining whether an image is a hyper-realistic recreation of a particular person or an amalgam derived from public internet photos can complicate enforcement and litigation, especially for emerging talent without robust representation.

Kusin emphasized that Kartel seeks explicit licensing agreements with models, arguing that doing so ensures compensation and consent. He noted, however, that not all companies will follow that model, and urged broader industry conversations about ethical safeguards and human participation in creative decision-making.

The technology itself continues to evolve. LoRA and similar adaptation methods allow firms to fine-tune generative models on a limited set of images to reproduce a consistent likeness across multiple outputs. Proponents say the tools can expand creative possibilities, reduce production time and increase accessibility for small brands. Critics point to the risk of job displacement, the amplification of unrealistic beauty standards, and the potential for deceptive or uncredited uses of a person's image.

Industry watchers say the next few years will be pivotal for how brands, creators and lawmakers respond. Some stakeholders are pushing for clearer contract language, model union discussions and regulatory frameworks to define consent, compensation and transparency for AI-generated media. Others caution that legal and policy changes typically lag behind rapid technological deployment, leaving individual creatives to negotiate protections in the interim.

For models like James, licensing a digital replica represents an attempt to capture value from emergent tools rather than cede control. "It's humans that do it," Kusin said of Kartel's process. "They just happen to be using AI tools, and we call them 10x creatives, because they move 10 times faster and have limitless creativity."

As fashion and advertising increasingly incorporate generative AI, the industry faces choices about how to balance cost and convenience with labor protections and creative integrity. The debate over whether AI will complement or largely replace human models is likely to continue as more brands test the technology, legislators consider new rules and courts interpret existing laws in the context of synthetic likenesses.