Levi’s made waves in 2023 when it partnered with Lalaland.ai to test AI-generated models for its e-commerce platform. The backlash was swift: accusations of eliminating diversity by replacing real models, criticism from labor advocates, and awkward public backpedaling. Three years later, AI-generated fashion models are everywhere — and the conversation around them has grown more nuanced than those early controversies suggested.
The technology has evolved from generating obviously synthetic faces to producing photorealistic human figures that are, in controlled testing, indistinguishable from photographs of real people. Platforms like PixelPanda now offer extensive libraries of AI avatars that brands can use for product marketing, virtual try-on imagery, and UGC-style content — without booking a single casting call. And unlike Levi’s clumsy first attempt, the current generation of tools was built with diversity as a core feature, not an afterthought.
But the rise of AI models raises questions that the industry is still working through, and the answers matter more than many brands realize.
The Scale of Adoption
It is difficult to get precise numbers because many brands do not disclose their use of AI-generated models. But the indicators are clear. Usage of AI model placement tools grew approximately 400% between 2024 and 2025, according to data from several platform providers who shared anonymized statistics with Clever Fashion Media. The fastest adoption is happening in three segments:
Fast fashion and ultra-fast fashion brands, which need to photograph hundreds of new styles weekly and cannot afford traditional model shoots for every item. Shein, Temu, and their competitors were early adopters, though none will speak publicly about the extent of their use.
Small to mid-sized DTC brands that previously could not afford model photography at all. For a brand that was using flat lays or mannequin shots, AI models represent an upgrade in visual quality, not a replacement of human workers.
Marketplace sellers on Amazon, Etsy, and Poshmark, where listings with model imagery consistently outperform product-only shots in click-through rate and conversion — but where the economics of hiring models for every listing make no sense.
The Diversity Argument Has Flipped
The initial criticism of AI models centered on diversity: by replacing real models, brands would reduce opportunities for models from underrepresented groups. This argument made sense in 2023. It makes less sense now, and here is why.
Traditional fashion photography has a well-documented diversity problem. Despite years of industry commitments, the Fall 2025 runway season saw only marginal improvements in racial, size, age, and disability representation. Model agencies still skew overwhelmingly toward specific body types and demographics. And for small brands, the economics of model casting often mean hiring whoever is locally available, which in many markets means limited diversity.
AI model libraries, by contrast, are designed from the ground up to represent a full range of ethnicities, body types, ages, and appearances. A brand using an AI avatar platform can show the same garment on models that reflect their actual customer base — a 55-year-old woman, a plus-size man, a person with a prosthetic limb — without the logistical complexity and cost of booking diverse casts for every shoot.
This does not eliminate the ethical concerns, but it reframes them. The question is no longer “AI models versus diverse real models.” For most small brands, it is “AI models versus no models at all” or “diverse AI models versus whichever one real model we can afford.”
What the Industry Actually Thinks
We spoke with professionals across the fashion ecosystem to gauge current sentiment. The responses were more varied than public discourse suggests.
Model agencies are unsurprisingly critical, though privately some agency executives acknowledge that the volume work — marketplace listings, catalog shots, social content — was never sustainable for their talent anyway. The real concern is about the ceiling: if AI handles 80% of model imagery, the remaining 20% of high-end editorial and campaign work becomes more competitive and potentially lower-paid.
Brand marketing directors are largely pragmatic. “We use AI models for our Shopify product grid and Amazon listings. We use real models for our seasonal campaign and social content,” one DTC brand founder told us. “It is not an either/or. We spend the same total budget but allocate it differently — less on routine catalog work, more on creative campaigns with real talent.”
Fashion photographers are divided. Commercial product photographers see the threat clearly and are pivoting toward editorial, campaign, and experiential work. Meanwhile, some photographers are actively embracing AI tools as part of their own workflow — using AI to generate concepts, composite elements, or extend sets.
Consumers appear largely indifferent, based on available data. A/B tests consistently show that shoppers engage with AI-generated model imagery at rates comparable to or slightly higher than traditional photography, provided the quality is good enough. Most consumers do not know — or particularly care — whether the model in a product listing is real.
The Ethics That Actually Matter
Much of the early ethical debate around AI models was focused on the wrong questions. The issues that deserve serious attention are:
Consent and likeness rights. Many AI model systems were trained on datasets that included real people’s faces and bodies, often without explicit consent. The legal framework around this is still evolving, but brands should be aware of the provenance of their AI models and favor platforms that use consented or synthetically generated training data.
Disclosure. Should brands be required to disclose when imagery features AI-generated models? There is no legal requirement in most jurisdictions, but some industry groups are developing voluntary standards. The argument for disclosure is transparency; the argument against is that it creates an artificial stigma that does not serve consumers.
Labor transition. The fashion modeling industry employs hundreds of thousands of people globally, many of them in economically precarious positions. As AI absorbs routine modeling work, the industry has a responsibility to think about transition support — though it is worth noting that the modeling industry itself has historically offered little job security or worker protection.
Body image. AI models can be generated to any specification, which creates both opportunity and risk. Brands can represent realistic body types — or they can generate idealized, unattainable physiques more efficiently than ever before. The technology is neutral; the choices brands make with it are not.
Where This Is Heading
The trajectory is clear: AI-generated models will become the default for routine commercial fashion imagery within the next two to three years. The technology is already good enough, the economics are too compelling, and consumer acceptance is established.
Real models will continue to play essential roles in high-end campaigns, editorial content, runway shows, and brand storytelling. The demand for exceptional human talent — models who bring genuine personality, cultural relevance, and creative collaboration to a shoot — will remain strong. What will diminish is demand for anonymous catalog modeling: the “stand here, hold this, turn left” work that comprises the majority of commercial modeling bookings.
For brands navigating this transition, a few practical principles apply:
Be deliberate about when AI serves your brand and when it does not. Not every image needs a real model, and not every image should use an AI one.
Invest the savings meaningfully. If AI reduces your photography costs by $50,000 per year, consider directing some of that toward better creative campaigns, higher-quality content, or — if you want to put your money where your values are — paying the real models you do hire more fairly.
Stay current. This technology is evolving in months, not years. The AI model imagery available today is substantially better than what was available six months ago, and six months from now it will be better still.
The fashion industry has always been about image. How that image is created is changing. What matters — authenticity, aspiration, representation, and the ability to help a customer imagine themselves in a garment — has not changed at all.