While doing my usual browsing on X recently, I came across a post accusing a certain TikTok account of being AI.
The account in question belongs to a model with over 50k followers and a bio that reads, “I find good Shein prices so u don’t have to. Plus size | Curvy” and a South African flag attached.
Prudence, as the model calls herself, has a discount code posted on her profile and a Shop My link where she curates the various looks she posts about on TikTok. As far as I can tell, none of her posts are videos, only stills. Her first post was in December 2025, and all she promotes is Shein.
Separately, one Vanelli Melli, a fashion influencer on TikTok, posted about her “AI twin”, which she came across when a friend sent her an AI fashion ad with a model that not only looks like her but also has the same style aesthetics.
“What do you mean? This is wild!” she says in the video post. Whether or not Prudence is AI, and it is genuinely difficult to tell, both cases point to something much larger already reshaping fashion at every level: the era of the AI model is not coming. It’s already here.
The industry is no longer merely experimenting with AI in image-making. It is integrating it into everyday production workflows.
Across fashion e-commerce, generative AI is increasingly being used to create marketing imagery more quickly and at significantly lower cost than traditional photo shoots. What began as testing has steadily moved toward operational use.
Retailers, including Mango, have already deployed generative AI imagery in commercial campaigns, using hybrid production methods that photograph real garments before digitally placing them onto computer-generated models.
The appeal is obvious. Campaigns can be produced faster, adapted for multiple markets simultaneously, and updated without reshooting entire collections.
Other brands, including H&M, have explored similar approaches through partnerships with AI modelling companies capable of generating diverse virtual figures for online retail imagery.
For e-commerce in particular, where brands require thousands of near-identical product images, AI offers a level of scalability traditional production simply cannot match.
Perhaps the most debated development, however, has come from brands experimenting with digital replicas of real models rather than entirely synthetic faces. In these cases, models collaborate with brands and agencies to create licensed digital versions of themselves that can appear in AI-generated campaigns. The promise is efficiency without erasing human participation. Models retain involvement in how their likeness is used, while brands gain flexibility in producing marketing content.
Even so, the implications extend far beyond efficiency. If a digital replica can appear in multiple campaigns without travel, styling teams, or studio crews, entire layers of fashion production inevitably shift. Photographers, stylists, makeup artists, lighting technicians, and the wider ecosystem surrounding a traditional shoot now sit within an industry confronting automation at scale.
And then there is the influencer economy, where Prudence and Vanelli Melli exist.
Here, AI becomes less institutional and far more ambiguous. Major brands operate within legal frameworks and public scrutiny. Anonymous or semi-anonymous social media accounts do not. An account promoting fast fashion through affiliate links does not need to disclose production methods unless platforms detect or enforce it.
If Prudence is AI, whether created by a brand, a marketing agency or an individual affiliate, she represents a different evolution altogether. AI is no longer simply a production shortcut but a persona designed to build trust, cultivate relatability and sell directly within communities built around authenticity and shared identity.
The plus-size and curvy fashion space online has historically functioned as a peer-driven environment shaped by real people sharing lived experiences alongside clothing recommendations. An AI presence operating within that space would blur the boundary between representation and simulation in ways audiences are only beginning to confront.
Vanelli Melli’s experience illustrates the opposite risk. As generative image tools improve, real creators increasingly encounter versions of themselves, including their faces, styling choices, or aesthetic signatures, reproduced without consent.
Social media platforms now require some level of disclosure for realistic AI-generated content, but enforcement remains uneven, and detection continues to lag behind creation.
What both stories ultimately reveal is that AI in fashion is no longer a speculative future. It is already embedded in how clothing is marketed, how visual culture is produced, and how influence itself can be manufactured, sometimes transparently and often not.
The question is no longer whether artificial intelligence will reshape fashion. It already has. The more urgent question is whether the industry, and the audiences it depends on, will still be able to recognise who, or what, they are looking at.
