Dior’s AI Dispute Shows Why Advertising’s Big Labels Fail

By Marcus Bennett

Three outlets had asked Dior for a response by Friday and received no answer to claims that its new animated holiday short “Night in Color” used generative AI, even though creative director Jonathan Anderson had told Women’s Wear Daily it was “all hand-drawn.”

The campaign’s most expensive marketing talking point was one extra finger. Last week, viewers noticed that the central cartoon character had one joint too many, along with warped lettering and a lazy eye. Online commenters concluded that generative AI made the film—a route many chief marketing officers boast about—and online outrage followed.

The film itself was terrific, which sharpened the credibility problem. Sylvain Chomet directed a talking-animal fairy tale about Christian Dior and Marlene Dietrich, and Dior released a making-of film to show off the craft.

Ben Wild, who runs a film studio with Netflix and Disney on its client list, said two people claiming to have worked on the film told him that AI had largely generated the motion and had generated the furniture and bees entirely.

Dior had confirmed none of Wild’s account but had not denied it either. Director Chomet denied using generative AI to make the film but acknowledged using it to smooth out the animation.

A spectrum, not a clean line

The quarrel around one film misses the normal operating model. Almost nothing people watch or read remains AI-free. Modern media production blends human and machine work across a spectrum, so the useful questions are where a project sits on that spectrum and whether the brand discloses its position.

• A phone blends several frames before the shutter button finishes its press.
• Photoshop can invent the other half of a sky.
• AutoTune has rescued pop vocals for more than 25 years.
• Video captions arrive automatically, while machine learning cleans dialogue in the edit suite.
• The debate’s originating column had every sentence read by Claude to check grammar and factual errors.

The clean divide between real and artificial ended around 2024. The gradient runs from light noise reduction to a creature imagined and drawn entirely by machine; anyone who insists on a clean line is selling a simplified label.

A generated bee probably will not change audiences’ reactions. A false claim of no AI will, especially when a brand sells hand-made creativity as part of its identity.

What does this change for luxury-brand operators?

For operators, the practical test is not whether a bee was generated. It is whether they can describe the workflow accurately and withstand scrutiny. A purity badge offers no protection when buyers can compare the claim with the credits, process and finished result.

In a feed crowded with machine video, visible and fallible human effort has become a luxury cue. Prada, Porsche, Levi’s and Apple all emphasized manual production for the same reason, and Dior wanted to join that hand-drawn circle. But a cue works only when it carries real time and cost.

Wild’s version is that Dior rushed into the film after learning that Prada had an animation legend on board, and that teams cannot accelerate this kind of work without AI. That account remains an allegation, not a confirmed production record.

If Dior had disclosed upfront that artists drew keyframes and machines filled the gaps, a few purists might have grumbled, but most people would still have admired the bees and accepted a modern brand. Instead, the absolute claim of pure manual labor became the ad’s main proposition—and its endemic flaw.

Why luxury carries more risk

A mass brand can survive a contradiction because nobody expects a Doritos ad drawn by monks.

Luxury sells provenance first and product second; for these brands, provenance is the product. A story about skilled hands that turns out to involve graphics cards does more than spoil one ad’s reception. It pushes buyers to reconsider the brand itself.

It is a shame, because Anderson deserved a better defense of a refreshing idea that produced lovely advertising. The larger lesson is that the AI-or-not argument has become advertising’s most pointless fight.

Brands gain little from boasting that work is AI-generated and little from claiming it is entirely handmade. Both positions grow harder to defend as production combines the two.

Brands do well to avoid the binary completely: the work is the work, and campaigns should stand on what they make, not on a label that no longer describes production.