Maggie Gyllenhaal Was Asked to Use AI to Recreate Marilyn Monroe. But it Flopped Compared to Dakota Johnson

← Back to the feed

Maggie Gyllenhaal Was Asked to Use AI to Recreate Marilyn Monroe. But it Flopped Compared to Dakota Johnson

Developing story first seen 5 hours ago

The Hollywood Reporter · 5 hours ago

Maggie Gyllenhaal has revealed that a fully AI-generated recreation of Marilyn Monroe was scrapped from her short film "Flesh Impact" because it lacked the emotional depth of Dakota Johnson's live-action performance. Speaking at a Venice Film Festival creativity summit, Gyllenhaal said producers had asked her to use licensable AI, trained on licensed Monroe footage, to overlay the star's face onto Johnson's; she ultimately discarded the AI version, calling it lifeless compared with what Johnson brought to the role. She also warned more broadly that AI-driven cost-cutting in filmmaking benefits the companies licensing the technology rather than filmmakers, and could make it harder for directors to secure traditional budgets.

The 17-minute film, made with the blessing of Monroe's estate holder Authentic Brands Group, stars Johnson as Monroe in her prime and Ellen Burstyn as an imagined older version of the star, alongside Peter Sarsgaard and Sepideh Moafi; it premieres in Venice's Sala Grande on 7 September. Gyllenhaal suggested a $40 million production could soon be made for $4 million, or even $40,000, using AI, questioning who would fund films "the old way" if cheaper alternatives sufficed. It was not her only headline moment at the festival: she also criticised the lack of female directors in this year's main competition, which includes only one woman-directed film among its original selections, down from six last year.

  • Gyllenhaal scrapped an AI-generated Marilyn Monroe for being less moving than Dakota Johnson's performance
  • She warned AI could undercut film budgets and who profits from cheaper production
  • Gyllenhaal also criticised Venice's main competition for lacking female directors

New here? Start with this

Maggie Gyllenhaal is an American actress and director whose short film, "Flesh Impact," is premiering at the 2026 Venice Film Festival. The film reimagines Marilyn Monroe, the classic Hollywood star who died in 1962, with Dakota Johnson playing her in her prime and Ellen Burstyn playing an older, imagined version of her; the production was made with permission from Authentic Brands Group, the company that holds rights to Monroe's image and estate.

The story centres on the growing use of artificial intelligence in filmmaking, particularly tools that can recreate the likeness of deceased celebrities using licensed footage. Gyllenhaal has spoken publicly about being asked to use such AI to digitally overlay Monroe's face onto Johnson's performance, raising questions about how this technology might affect acting, budgets and who profits from it.

This matters because it touches on a wider debate within the film industry about AI's role in production costs, creative authenticity and the livelihoods of actors and filmmakers, at a time when AI tools are becoming increasingly capable and accessible. The Venice Film Festival itself is one of the most prominent events in international cinema, giving such comments significant visibility.

Both sides, in good faith

The strongest fair case each way — we don't pick a winner.

The case for

Advocates of AI-assisted filmmaking argue that licensed, consent-based tools like the one trialled here could democratise the industry, letting talented directors tell ambitious stories without needing tens of millions of pounds in backing. They point out that Authentic Brands Group's involvement shows estates can be properly compensated and protected rather than exploited, and that testing such technology in good faith, as Gyllenhaal did, is how the industry responsibly learns its limits. For them, a tool that can lower costs is not inherently a threat to artistry; it simply widens who gets to make films, and filmmakers remain free to reject it, as happened here, when it falls short creatively.

The case against

Sceptics, echoing Gyllenhaal's own misgivings, argue that AI-driven cost-cutting mainly enriches the companies that license the technology rather than the directors, actors and crews who make films work, and could erode the traditional budgets and jobs that sustain the craft. They worry that even consensual, licensed recreations of a performer's likeness set a precedent that normalises replacing human performance with synthetic imitation, undervaluing the emotional depth and lived judgement an actor brings, as this very experiment demonstrated. Combined with concerns about shrinking opportunities for women directors, they see this moment as evidence that unchecked technological shortcuts risk narrowing, not expanding, who gets to make meaningful cinema and how well it is rewarded.

More coverage

AI Government Politics Research Science Technology

Read the full article at the source →