Report: Disney’s Attempts to Experiment With Generative AI Have Already Hit Major Hurdles

As Silicon Valley has pushed the world more and more into trying to make the generative AI boom sustain itself, Hollywood is still standing on the precipice of a transformative moment. Studios are grappling with the purported potential (and demands for cost savings) artificial intelligence models may bring, weighed against the legal minefields exploiting such technologies can represent—and an increasing public backlash to the technology.

Disney is certainly no exception, as the company is already familiar with both the legal headaches and the PR nightmare generative AI can represent. But a new report from the Wall Street Journal claims that there’ve been even more attempts behind the scenes at Disney’s studio to try and utilize generative AI technologies… neither of which purportedly went very far or well, for very different reasons.

Two upcoming productions that tried to navigate potential use of generative AI mentioned in the WSJ report are the upcoming live-action Moana remake and Tron: Ares. For the former, Disney reportedly planned to work with an AI company called Metaphysic to create a digital deepfake of actor Dwayne Johnson, set to reprise his role as the demigod Maui in the remake. In an attempt to reduce the number of days Johnson would be required on set for production, the alleged plan was to have Johnson’s cousin, Tanoai Reed, act as a stand-in who would have Johnson’s deepfaked face put over his performance in post-production.

Although the plan was for a “small number of shots,” according to WSJ’s report, after 18 months of negotiation and work between Disney and Metaphysic, none of the shots using Reed’s performance will be in the final movie when it releases in July 2026. WSJ’s report cited concerns over data security on Disney’s end, as well as the legal question that lingers over any broader embrace of generative AI technology in Hollywood: who, exactly, owns the end product when generative AI models are used to create even a part of it?

That thorny question of ownership has already seen Disney take legal action against AI companies over claims of illegal misuse of copyrighted material to train their models. In June this year, Disney teamed up with Universal to sue Midjourney over what the suit described as a “bottomless pit of plagiarism,” accusing the AI company’s image generator of breaching copyright laws to distribute and create images trained on the studios’ library of characters and franchises.

But copyright is not the only concern Disney faces when it comes to ideas around generative AI: the studio is also increasingly navigating potential publicity nightmares as social backlash to the use of the technology increases.

In another example in WSJ’s report, it’s alleged that Disney executives pitched creatives on the set of Tron: Ares on including a generative AI character in the film, which itself is already about artificial intelligences escaping the digital world of “The Grid” to be exploited as military contractors in the real world. According to WSJ, the character would’ve been called “Bit” and acted as a potential companion to Jeff Bridges’ returning Kevin Flynn, and built off of context provided by a writer, the generated character would then be recorded and deliver lines performed by an actor, responding as if the model itself were Bit.

The report claims that similarly the idea was stymied again by legal discussions at the time, amid negotiations with unions, as well as the fact that Disney executives were purportedly told to drop the idea internally because “the company couldn’t risk the bad publicity.”

Disney is, of course, no stranger to public embarrassment when it comes to its properties and AI, either. Marvel was lambasted for the use of generative AI to create the opening title sequence to its Disney+ series Secret Invasion in 2023, and found itself defending itself from accusations of its use once more for the early marketing campaign for Fantastic Four: First Steps. Earlier this summer, Disney’s investment into Epic Games was touted through the arrival of a generative-AI-enhanced Darth Vader avatar in Fortnite to promote the battle royale video game’s then-ongoing Star Wars event, “Galactic Battle”, where Darth Vader could be recruited by players, using a deepfake model of the late James Earl Jones’ voice to interact with players in real time.

Players promptly figured out ways to get around Epic’s content restrictions and get the generative Vader to swear and use slurs. Although Epic managed to fix the bugs within 30 minutes of the Vader character’s appearance going live in Fortnite, several videos of the exploits went viral on social media. SAG-AFTRA also filed an unfair labor practice charge against Epic over the use of generative AI denying a human actor the chance to voice the role (Jones’ estate had already sold the rights to his voice to the Ukrainian tech company Respeecher in 2022 before his passing), but rumors recently swirled that the union dropped the charges in the wake of signing a new contract last month.

The legal outcome of Disney and Universal’s lawsuit is still to be decided, but what is seemingly clear is that the potential AI takeover of Hollywood that has been feared with the proliferation of generative AI may not be as close as some people (and some companies) expect.

Want more io9 news? Check out when to expect the latest Marvel, Star Wars, and Star Trek releases, what’s next for the DC Universe on film and TV, and everything you need to know about the future of Doctor Who.