For most of fashion’s modern history, the face selling a dress or a pair of jeans belonged to a real person who turned up to a studio, stood under the lights and was paid for the day. That arrangement is quietly being rewritten. A growing number of retailers are generating their campaign imagery with artificial intelligence, producing polished “models” who never existed, wearing clothes that were often never physically photographed on a human body.
The trend was explored in a recent feature by Inside Retail Australia, which charted how the technology has shifted from a curiosity into something resembling standard practice for a slice of the industry. What was once a gimmick trotted out for a single experimental campaign is increasingly being folded into the everyday production line of e-commerce imagery.
From novelty to normal
The direction of travel has been obvious for a couple of years now. Spanish giant Mango launched a teen campaign built entirely around AI-generated models in 2024, describing it as part of a broader push to lean into the technology across its business. Levi Strauss drew attention earlier when it partnered with a digital studio to supplement its human models with AI-generated ones, framing the move as a way to show garments on a wider range of body types and skin tones.
The commercial logic is not hard to follow. A traditional shoot involves models, photographers, stylists, hair and makeup, location hire, travel and post-production, and it has to be repeated every time a new range drops. Generative tools promise to collapse much of that into a software subscription and a few hours of prompting, letting a brand produce hundreds of on-model images for a new season without booking a single studio. For fast-moving online retailers pumping out thousands of product listings, the appeal is speed and scale as much as cost.
The quality has also crept up to the point where the uncanny stiffness of early attempts is harder to spot. Where AI faces once betrayed themselves with melted hands or glassy eyes, the newer systems produce imagery that many shoppers scroll straight past without a second thought, which is precisely what worries the critics.
Two camps form
Supporters inside retail argue the technology democratises something that was previously expensive and exclusionary. Smaller labels that could never afford a professional shoot can suddenly present their clothes with the gloss of a major brand. Advocates also point to the inventory problem that generative imagery solves: a garment can be shown on dozens of synthetic body shapes, ages and complexions without the logistical nightmare of casting that many real people, which some argue is a genuine step toward representation rather than away from it.
The opposing camp sees something more troubling. Modelling agencies and the people they represent view the shift as straightforward job displacement, with human work being automated out of a creative industry that already offers precious little stability. There is also the authenticity question. Consumer research consistently shows that trust frays when audiences feel they have been deceived, and a synthetic model wearing clothes that no human has ever tried on tells a shopper nothing about how a garment actually sits, drapes or fits. Critics warn that the practice risks deepening the gap between the airbrushed image online and the reality that arrives in the delivery satchel.
Then there is the body-image dimension. An industry long criticised for promoting unrealistic standards is now capable of generating a flawless human who is not constrained by biology at all, free of pores, asymmetry or the limits of a real body. Mental-health advocates have flagged the obvious hazard in handing that capability to marketers whose job is to make clothes look as desirable as possible.
What it means for Australia
Australia’s fashion sector is smaller than the European and American markets driving this trend, but it is far from insulated. The local industry leans heavily on imported e-commerce playbooks, and global platforms that adopt AG imagery at scale set the template that Australian brands tend to follow. For a domestic label competing against international fast-fashion sites on price and turnaround, the temptation to cut production costs with synthetic imagery will be real.
The Australian Fashion Council has spent recent years pushing to professionalise and protect a local industry that supports tens of thousands of jobs, from designers and machinists through to the models, photographers and stylists who make campaigns happen. A rapid shift to generated imagery would hit those creative roles first, in a country where that kind of freelance work is already precarious.
There is a regulatory angle too. Australia does not yet have dedicated rules forcing brands to disclose when an image is AI-generated, and consumer law around misleading conduct was not written with synthetic models in mind. The Australian Competition and Consumer Commission has shown appetite for pursuing misleading online conduct, and it is not difficult to imagine a future complaint arguing that a shopper was misled about a product by imagery that never involved the actual garment. Industry bodies overseas have begun calling for mandatory labelling, and Australian regulators will eventually face the same question.
What is next
The near-term likelihood is a split market rather than a clean winner. Volume-driven online retailers will keep adopting AI imagery wherever it saves money and no one complains, while premium and heritage brands increasingly make a selling point of the opposite, marketing the human photographer, the real model and the genuine shoot as a mark of authenticity worth paying for. Luxury houses have already signalled that human craft is part of what they are selling, and some are betting that the backlash against synthetic content becomes a point of difference.
For Australian retailers, the practical questions are arriving fast. Do you disclose when an image is generated, and does disclosure dent sales or build trust? Do the savings survive once you factor in licensing, the risk of consumer complaints and the reputational cost of being caught out? And what happens to the local creative workforce if the imagery that once sustained it can be conjured from a text prompt? The technology has clearly crossed from experiment into tool. How Australia chooses to label it, regulate it and value the human work it replaces is the part still being decided.
Sources: Inside Retail Australia.
