Vogue's AI Ad Uproar: Fashion Industry Grapples with Digital Models

The fashion industry is grappling with a growing wave of concern as artificial intelligence increasingly influences its visual landscape. This unease, initially sparked by brands like Levi’s utilizing AI-generated models for “diverse” ad campaigns, recently intensified following the appearance of an AI model in a Guess advertisement featured in Vogue’s July print edition.

For Sarah Murray, a commercial model, the sight of an AI-generated model in a 2023 Levi’s ad evoked feelings of sadness and exhaustion. Levi’s had partnered with AI studio Lalaland.ai to create digital models for more inclusive advertising, a move that drew sharp criticism, with one publication labeling it “artificial diversity.” Murray articulated a common sentiment among human models: “Modeling as a profession is already challenging enough without having to compete with now new digital standards of perfection that can be achieved with AI.”

Two years later, these anxieties have escalated. The recent Vogue ad for Guess showcased an AI-generated model embodying traditional North American beauty standards: thin yet voluptuous, with glossy blonde hair and pouty lips. The internet quickly erupted, largely due to Vogue’s status as a fashion arbiter. While Vogue clarified the image was an advertisement, not editorial content, and met its advertising standards, many in the industry saw this distinction as negligible.

This incident has amplified questions about the role of humans in creative industries touched by AI. When high-quality visual content can be produced by AI at a fraction of the time and cost, what becomes of human models, photographers, stylists, and set designers?

The Economic Imperative for AI

Cost is a primary driver behind brands’ adoption of AI models. Sinead Bovell, a model and founder of the WAYE organization, who wrote about CGI models five years ago, identifies “e-commerce models” as the most vulnerable to automation. These models typically pose for online advertisements and product displays, forming the financial backbone for many in the profession, distinct from the often unattainable looks of high-fashion editorial models. “E-commerce is where most models make their bread and butter,” Bovell noted.

Paul Mouginot, an art technologist who has collaborated with luxury brands, explained that working with live models, especially for extensive product shoots, is simply expensive. AI now enables brands to start with a flat-lay product image, place it on a photorealistic virtual model, and even position that model in a coherent setting, generating images that resemble genuine fashion editorials. This practice isn’t entirely new; companies like French retailer Veepee have used virtual mannequins since at least 2013, and other major brands such as H&M, Mango, and Calvin Klein have also incorporated AI models.

Fashion writer Amy Odell put it plainly: “It’s just so much cheaper for [brands] to use AI models now. Brands need a lot of content, and it just adds up. So if they can save money on their print ad or their TikTok feed, they will.”

PJ Pereira, co-founder of AI ad firm Silverside AI, emphasized the issue of scale. He noted that the traditional marketing system was designed for brands to produce only a handful of major content pieces annually. Social media and e-commerce now demand hundreds or even thousands of content pieces, a volume too expensive for most brands, particularly smaller ones, to maintain with traditional methods. “There’s no way to scale from four to 400 or 400,000 with just process tweaks,” Pereira stated, suggesting that the shift is not primarily about undercutting artists but about developing a new system to meet demand.

Concerns Over “Artificial Diversity” and Model Rights

Despite the economic advantages, models like Sarah Murray express deep reservations. She challenges the notion that AI merely “supplements” human talent, arguing that countless human models would eagerly fill the roles AI now occupies. Murray believes this shift disproportionately impacts “non-traditional,” or diverse, commercial models, citing the Levi’s ad as an example where diversity was artificially generated rather than genuinely hired.

Bovell terms this phenomenon “robot cultural appropriation,” where brands generate specific identities, particularly diverse ones, for storytelling without the technology’s creators sharing those identities. While Pereira suggests it’s impractical to photograph every garment on every body type, this perspective has not allayed the fears of diverse models. Murray highlighted a growing concern: “We already see an unprecedented use of certain terms in our contracts that we worry indicate that we are possibly signing away our rights for a brand to use our face and anything recognizable as ourselves to train their future AI systems.”

To address these concerns, former model Sara Ziff, founder of the Model Alliance, is advocating for the Fashion Workers Act. This proposed legislation would require brands to obtain clear consent and provide compensation for using models’ digital replicas. Mouginot acknowledged that digital replicas could allow models to participate in multiple shoots simultaneously and generate additional income, which is “precious when a sought-after model is already traveling constantly.” However, he also cautioned that “what few players gain can mean fewer opportunities for many others.”

Bovell advises models to adapt by building personal brands, differentiating themselves, and exploring new revenue streams like podcasting or brand endorsements. “Start to take those opportunities to tell your unique human story,” she urged, “AI will never have a unique human story.”

Niche Applications and the Fight Against Homogeneity

While some platforms may eventually abandon human models, Mouginot believes there remains a human desire for the “sensual reality of objects, for a touch of imperfection and for human connection.” He noted that many successful models are celebrated for distinctive, slightly imperfect traits that are inherently charming and difficult to replicate with perfect digital precision.

Creative studio Artcare aims to bridge this gap. CEO and co-founder Sandrine Decorde describes her team as “AI artisans” who use tools like Flux from Black Forest Labs to fine-tune AI-generated models, imbuing them with a touch of unique humanity. A significant portion of Artcare’s work involves producing AI-generated babies and children for brands. Decorde argues that this application makes ethical sense, given the historical complexities of employing minors in fashion and the high market demand.

Decorde also pointed out the homogeneity often seen in AI-generated models, such as those from Seraphinne Vallora, the agency behind Vogue’s Guess ad. She observed that their lips are frequently too perfect and symmetrical, and their jawlines identical. “Imagery needs to make an impact,” Decorde said, emphasizing that a model embodies a brand, and distinctiveness is key. Pereira’s firm actively combats AI homogeneity through intentional training, warning that without careful prompting and diverse datasets, AI will amplify existing biases.

An Uncertain but Inevitable Future

Claudia Wagner, founder of modeling booking platform Ubooker, views the current use of AI modeling technology in fashion as largely experimental. While acknowledging the technical interest of the Guess ad, she found it neither impactful nor groundbreaking. “It feels like another example of a brand using AI to be part of the current narrative,” Wagner commented. “The real value will come when it’s used with purpose, not just for visibility.”

Brands are undoubtedly gaining visibility from AI use. Pereira shared an example of a fully AI-generated product video on TikTok that garnered over a million views, mostly negative comments, yet resulted in significant engagement and a steep hike in sales. He, like Wagner, believes AI models are here to stay, with AI processes becoming integrated into creative workflows.

Pereira noted varying levels of comfort among brands: some embrace fully artificial models, others prefer licensing real people’s likenesses for synthetic shoots, and some avoid it entirely due to audience acceptance concerns. Wagner underscored that human talent remains central, especially when authenticity and identity are integral to a brand’s story. This is particularly true for luxury heritage brands, which typically adopt new technologies slowly.

While many high-fashion brands are quietly experimenting with AI, Mouginot observed that many are still defining their AI policies and avoiding fully AI-generated people for now. This context made Vogue’s inclusion of an AI model particularly surprising. Bovell pondered whether the ad was Vogue’s way of gauging public reaction to the merger of high fashion and AI. The initial response has been largely negative, leaving it unclear how the magazine will navigate the backlash. As Odell concluded, “What Vogue does matters. If Vogue ends up doing editorials with AI models, I think that’s going to make it okay. In the same way the industry was really resistant to Kim Kardashian and then Vogue featured her. Then it was okay.”