AI images instead of fashion photography

Following Zalando and H&M, Zara will use AI-generated images with digital clones of models instead of photoshoots.

listen Print view

(Image: Praswin Prakashan /)

2 min. read

The fast-fashion brand Zara, part of the Spanish Inditex group, is heavily investing in AI-assisted image production. The fashion group intends to vary photos of real models using generative software, rather than conducting a new photoshoot for each product variant. Numerous variations are to be created from photos of models taken once, using AI, for example, with different colors, cuts, or accessories.

The pose, body shape, and face of the real models will be preserved, while the clothing and backgrounds will be exchanged. This is intended to drastically reduce the production time and costs for new product images for the webshop and app. Furthermore, it offers the opportunity to ride the fast-moving fashion wave in real-time.

Videos by heise

Zara is not the first relevant company to increasingly rely on AI. Several months ago, H&M announced its intention to work more extensively with "digital twins," AI-generated images of its models, in the future. The models will retain full rights to their clones, and compensation will largely correspond to conventional fees. Zalando also announced in May that it would meet the rapid pace of fashion change with AI support.

The companies emphasize that the "clone" strategy is intended to complement human work, not replace it entirely. Critics, including the British Association of Photographers, see it differently. There are fears that the use of generative AI will reduce the number of traditional assignments for photographers, models, and production teams. The consequence could be a gradual decline in assignments, particularly affecting younger and freelance creatives. It is also unclear to what extent a transparent labeling requirement for AI-generated images will be regulated and enforced in the future.

(swi)

Don't miss any news – follow us on Facebook, LinkedIn or Mastodon.

This article was originally published in German. It was translated with technical assistance and editorially reviewed before publication.