The Fashion Fiction We’re Willing to Buy

When MANGO launched its first fully AI-generated ad campaign last year, it made headlines: one of the first major fashion brands to swap models and sets for pixels and prompts. But now it’s gone a step further.

This time, it’s not just campaign visuals or aspirational lookbooks. MANGO has begun generating actual Product Detail Page (PDP) images using AI , the main visuals that shoppers rely on to decide whether to click Add to Cart.

This shift is seismic. It means the brand no longer needs studios, photographers, stylists, makeup artists, models, catering, travel, insurance, licensing fees , all the real-world overhead that’s sustained entire ecosystems of creative jobs for decades. For a company selling clothes for 30 euros a piece, the math is simple: AI wipes out creative costs by 60 to 80 percent, accelerates production overnight, and spits out as many variations as you want ( think localization).

MANGO’s move isn’t revolutionary in principle , IKEA’s catalogs were filled with 3D-generated furniture 10 years ago, long before AI came along. What’s new is the accessibility, scale, and casual acceptance of fully synthetic visuals at the final moment of conversion. The idea that a pair of 15 euro shorts deserves the same simulation treatment as a wooden kitchen chair shows just how deep the trust in plausible fictions now runs.

Of course, that efficiency has a cost and not just for the people now wondering what happens to their careers. It also leaves us with a deeper, stranger question: What exactly are we looking at when we browse these pages? And what, then, are we buying?

“What you see on MANGO’s product page isn’t the garment, not exactly. It’s closer to an artist’s interpretation, only the artist is an algorithm. This digital stand-in is built from patterns, textures, and styling cues the AI model has learned from thousands of examples of how fabric should drape and light should fall. It’s not reality captured but reality approximated: a synthetic vision of the dress as it should look when worn, not as it ever actually did. The folds, the silhouette, the casual bend of an elbow, all are persuasive illusions crafted to sell a promise that the physical garment will live up to this virtual ideal.”

So, when we say the difference between AI imagery and light-based photography is that one is rooted in reality and the other is not, well, MANGO’s experiment shows that the truth is more complicated.

In traditional studio photography, light bounces off a real dress, on a real body, in a real moment. There are flaws, wrinkles, and imperfect poses; the physical object becomes an image by passing through the lens. But in these AI-generated PDPs, there is no lens. There is only a prompt, a data model, and a convincing illusion.

And yet, the final result can’t drift too far from the physical world. The skirt must arrive looking like the image, or the promise, and the sale collapses. In a way, the AI’s invented image is a fiction with very strict rules: it must stay loyal to a reality that it has never actually seen.

So the question becomes: Is the AI rendering less real because it was never photographed? Or is it, perversely, more real because it delivers a version of the product closer to the brand’s ideal?

This is the paradox MANGO has put on the table for the entire retail industry. It’s not just about how many jobs will disappear. It’s about how far we’re willing to go in trading captured reality for a manufactured approximation, so long as the package that lands on our doorstep doesn’t break the illusion.

Maybe the product photograph was never just about truth. Maybe it was always about trust. And maybe, in the era of AI-generated shopping, that’s the only thing that really needs to stay real.