Published October 21, 2024 · By CampaignsLive · Industry
Fashion editorial photography is one of the higher-equity, lower-volume corners of the commercial image market. A Vogue editorial is the work of named photographers, named stylists, named makeup artists, named models, all of whom are credited and many of whom are central to the magazine’s identity. The work is, in a particular sense, the opposite of the e-commerce product photography that Mango’s Sunset Dream campaign tested in mid-2024. The economic argument for AI in editorial photography is weaker; the cultural argument against it is stronger; the disclosure norms are tighter.
The way generative AI moved into and around fashion editorial in 2023 and 2024 is a useful study in how a high-equity creative category negotiates a technological shift it cannot ignore. The boundaries were not set by formal policy. They were set by what publications, photographers, and brands tried, and by what the audience, the talent, and the broader industry tolerated.
Where AI showed up in fashion editorial
Three places, with three different reception patterns.
As a stated creative concept. Vogue and several of its sister publications ran editorial features that foregrounded AI as the creative idea — features about AI in fashion, features built around AI-generated reference material in collaboration with named photographers, features where the AI was a credited creative collaborator in a way structurally similar to how Heinz had treated DALL·E in “Draw Ketchup” in 2022. These features attracted attention, were credited cleanly, and did not produce significant industry friction.
As an undisclosed production tool. A different and more controversial category. Editorial work that used AI in compositing, environmental extension, or sky and background replacement, without disclosing the use to readers. This category was harder to detect from outside the production and consequently harder to comment on definitively. The publications that ran in this register did not, generally, advertise it. The category exists; the boundary of acceptable use within it is still being negotiated.
As a brand-side commission. Fashion houses commissioning AI-generated imagery for in-magazine advertising — full-page placements, gatefold inserts, sponsored content adjacent to editorial. This was the most visible category, the most economically meaningful, and the most contested. The boundary between editorial integrity (the magazine’s responsibility) and advertising creative (the brand’s responsibility) became the active site of disagreement.
The Calvin Klein moment, and what it set up
In late 2023 and through 2024, Calvin Klein ran a sequence of campaigns that became the test case for the third category. The brand commissioned imagery that ran in major fashion publications, using a combination of traditional photography and AI-augmented production. The exact provenance of specific images was, by brand decision, not always disclosed. The publication side accepted the work and ran it; the editorial side maintained that the commercial pages were the brand’s responsibility, not the publication’s.
The case set up a pattern that has held through 2025. Major fashion publications have, broadly, taken the position that brand-side disclosure is the brand’s call within paid placements, but that the editorial pages remain governed by editorial standards that require disclosure of AI provenance. The line is not airtight — there have been edge cases in both directions — but it has been the working norm.
The brands that have run the most AI in their fashion advertising creative have, with some exceptions, followed the pattern Mango followed in its e-commerce campaigns: the AI is in the production stack, the production stack is the brand’s business, the audience-facing framing is honest about the broad shift without itemizing every campaign. This is a working compromise that has held longer than the more strident positions on either side.
Why fashion editorial is harder than fashion commercial
Three reasons.
Talent identity matters more. A fashion editorial is, in part, a vehicle for celebrating the work of named contributors — the photographer, the stylist, the model. AI-generated imagery has a structural problem with this kind of celebration. There is no named photographer behind an AI-generated cover; the credit is fragmented and uncomfortable. The publications that have tried to navigate this — by crediting the AI tool, the prompt engineer, or the team that ran the production — have produced credit structures that fashion’s traditional contributor model does not absorb cleanly.
The audience attends differently. A reader of fashion editorial is, on the whole, looking more carefully at the image than a reader of paid social or an e-commerce thumbnail. The seams that are invisible at thumbnail scale become visible at full-page editorial scale. The bar is higher, the audience is more attentive, and the failure modes are more consequential.
The labor norms are stronger. Fashion editorial production employs a specific set of crafts at a specific level of professional organization. The photographers, stylists, and models who work at the editorial level operate inside a labor structure that is more visible than the e-commerce production world. A high-profile editorial that displaces this work attracts a different and louder set of responses than an e-commerce campaign that does the same.
Where the boundary has actually settled
By the end of 2024, the working pattern across major fashion publications looked like this.
Editorial pages. AI used only as a stated creative concept, with clear credit and disclosure. Augmentation of traditional photography (background extension, color work, sky replacement) is acceptable in proportion and where disclosed.
Commercial pages. Brand decision, with publication minimum-disclosure standards. The brand absorbs the audience response; the publication maintains the right to refuse work that violates emerging norms.
Online supplementary content. A grayer area, with looser norms. AI work that would not run in print editorial has appeared in online editorial, where the labels are sometimes less visible and the production economies favor faster, looser content production.
The pattern is more nuanced than the categorical “AI is or is not allowed in fashion photography” framing the trade press has periodically reached for. The boundary is not a line; it is a set of overlapping norms that vary by surface, by brand, by publication, by time.
What this leaves for the next few years
The unfinished question is what happens as the technology continues to improve. The current boundaries depend, in part, on the audience’s ability to recognize AI-derived imagery. As that ability shrinks — as the seams become harder to find — the bright-line cases get more pressure. The editorial that ran AI-augmented work in 2024 without disclosing it was a smaller violation than the editorial that runs the same work in 2026 without disclosing it, because the gap between disclosed and undisclosed has widened.
The publications that have done well navigating this have been the ones that committed to disclosure norms early and held them as the technology improved. The publications that have left themselves the most room have been the ones that did not.
For the related discussion of disclosure norms in commercial production, see Marvel’s Secret Invasion AI Title Sequence. For the e-commerce parallel, see What Mango’s All-AI Catalog Reveals.