A new batch of AI-generated short films is drawing serious heat for animating photos from a 1976 erotic magazine without the subjects’ consent. According to Futurism AI, the collection, titled “Sh(AI)ved,” was produced by Norwegian company Multiformat and distributed by streamer Cultpix. It premiered in Cannes this week, though notably not at the official Cannes Film Festival.
The project uses generative AI to turn still images of the original magazine’s models into moving footage, complete with sound and dialogue. Critics on Bluesky and X are calling it nonconsensual porn, pointing out that many of the women photographed half a century ago have likely died. What stands out here is that the filmmakers don’t seem to have asked whether any of the subjects were contacted, or whether their estates were.
What Cultpix is saying
Cultpix CEO Rickard Gramfors framed the project as a conversation starter about shifting cultural norms. “We want to use the latest technology to stimulate a discussion about attitudes to images that are now half a century old,” he told Variety, as reported by Futurism AI. “What was once considered shocking ‘adult’ material now seems remarkably innocent by today’s standards.”
The company’s official X account then jumped into the comment section to defend the work, calling it “an experiment.” In a second post, Cultpix argued consent wasn’t an issue because the original performers signed off in the 1970s. “These were paid [performers] who consented to have their hanky panky recorded. We added motion and sound,” the account wrote.
That argument is the part legal scholars will likely tear apart first. Consent to be photographed in 1976 isn’t the same as consent to have your likeness animated, voiced, and redistributed in 2026 by a technology that didn’t exist when you signed the release.
Why this matters for the industry
The “Sh(AI)ved” backlash lands in the middle of an already ugly year for AI and nonconsensual imagery. Futurism AI notes that Elon Musk’s Grok chatbot has been caught generating images that digitally “undress” women and children, fueling broader regulatory pressure on generative tools.
A few takeaways for anyone building or shipping AI media products:
- Historical consent isn’t a free pass. Repurposing old releases for modern synthetic media is going to invite lawsuits and platform takedowns.
- Estates are the next legal frontier. Posthumous likeness rights vary wildly by jurisdiction, and France, where the films premiered, has some of the strongest protections in the world.
- “It’s an experiment” won’t fly as a defense. Studios shipping AI projects need real consent frameworks, not artistic justifications.
- Distribution platforms carry risk too. Streamers and festivals that host this kind of content are now part of the liability chain.
What comes next
Expect more pressure on platforms hosting AI-resurrected likenesses, especially of deceased performers. Industry groups have been pushing for clearer rules around synthetic media, and a case this public, with this much social media outrage, gives regulators an easy example to point at.
For practitioners, the lesson is simple: if your AI pipeline touches a real person’s face or body, build the consent layer before you build the model. Retrofitting ethics after a viral backlash doesn’t work.
Full reporting is available at the original Futurism AI source.