About You Retires Photography: What the Celebrated Case Really Shows

A case currently making the rounds through every retail and fashion trade outlet: About You is closing its classic photo studio in October and will have product images generated entirely by AI going forward. Cost per item is said to drop from up to 80 euros to about one euro, which, at up to 300,000 products photographed per year, adds up to tens of millions in savings. Revenue on AI-generated product images averaged 9.2 percent higher than on classic shots in testing. The in-house technology is now also being sold to other brands under its own name, including Betty Barclay, S.Oliver, and FC Bayern München.

As someone who has co-run a studio offering photography and CGI side by side for years, I read these numbers differently than most of the current headlines suggest.

The 80-euro figure is the actual sleight of hand

The most striking number in the case is the jump from 80 to 1 euro. What's almost never included: what exactly did those 80 euros buy? Classic fashion photography at that price point typically includes not just pressing the shutter, but styling, a model, lighting setup, art direction, and retouching – a production process that makes decisions, not just delivers images. A one-euro AI-generated image doesn't replace that process, it skips it. That's not an unfair comparison because AI is worse – it's an unfair comparison because two different products are being pitted against each other as if they were the same thing.

The comparison limps for a second reason too, one I know from our own production: nobody pays 80 euros per image in a well-run process. At Gieske Studios, we produce complete image sets of up to six views per item – model, styling, retouching, everything included – for around 50 euros per set, not per single image. The workflow is timed so images are ready to upload the next day. If About You's internal calculation genuinely landed on 80 euros per image, that says less about the limits of photography than about the efficiency of its own previous process. A comparison between AI and a well-organized photo studio would likely look quite different from one between AI and an expensive single-image workflow.

What the 9.2 percent revenue increase could actually mean

A higher revenue figure for AI images sounds like clear-cut proof. But it can just as easily be read differently: if a large share of the previous 300,000 catalog images were produced on a tight budget with little time per item – plausible at that volume – then the test may not be measuring "AI beats photography" so much as "more deliberate, more consistent visual language beats half-hearted mass production." That's exactly what I described in the comparison between CGI and photography: at high volumes, the technique that delivers consistency across hundreds of variants almost always wins – regardless of whether that technique is called CGI or generative AI. The case may well be demonstrating that effect rather than a fundamental AI advantage.

The point the coverage leaves out

One detail only shows up at the margins in the reporting so far: AI models look realistic but aren't human, and the uncanny valley lurks exactly where fashion is hardest to simulate – skin, hands, and above all the way fabric falls. How a garment drapes over a real body in motion is physically one of the most complex problems in all of CGI and rendering practice. Anyone working in this field knows: that's precisely where reliably sellable production separates from work that's technically impressive but unconvincing in the details. Whether AI-generated models clear that bar equally well across all product categories, not just the ones tested so far, is an open question that a 9.2 percent average revenue figure doesn't answer on its own.

The labeling requirement no report mentions

Notably, none of the articles on the About You case mention that starting August 2, 2026 – just days after the celebrated rollout – Article 50 of the EU AI Act takes effect. Photorealistic AI images that depict products true to life can fall under it, even without any real person being shown. A fashion retailer switching its entire image catalog to AI should really have settled that question before retiring the studio – not once the first fines arrive.

What's genuinely real about the case

None of this means the case is wrong or dishonest. For catalog goods at very high volume, low emotional stakes, and frequent turnover – exactly the segment About You operates in – the math is probably genuinely positive, for the same reasons CGI has long had the edge there. The mistake isn't in the result, it's in drawing a general conclusion about "photography versus AI" from a case built on fast-fashion catalog goods. For brands whose value comes precisely from people, emotion, and hand-crafted distinctiveness, a different calculation applies – one that dissolves exactly in averages taken across 300,000 items.

Conclusion

The About You case is currently being treated as proof that classic product photography is becoming obsolete. Read more closely, it mainly shows: for catalog goods at this scale, the old approach to photography was already the wrong one – not because photography is inherently inferior, but because this segment was always built for systematized image production. The real question no article has asked yet isn't "photography or AI," but which product segment deserves which visual language – a distinction that should come before any cost calculation, not after.