Is it still Disney magic if it's AI?

Mint

5 days ago

When Disney began working on a new, live-action version of its hit cartoon 'Moana," executives started to ponder whether they should clone its star, Dwayne Johnson.
The actor was reprising his role in the movie as Maui, a barrel-chested demigod, but for certain days on set, Disney had a plan in place that wouldn't require Johnson to be there at all.
Under the plan they devised, Johnson's similarly buff cousin Tanoai Reed—who is 6-foot-3 and 250 pounds—would fill in as a body double for a small number of shots. Disney would work with AI company Metaphysic to create deepfakes of Johnson's face that could be layered on top of Reed's performance in the footage—a 'digital double" that effectively allowed Johnson to be in two places at once.
What happened next was evidence that Hollywood's much-discussed, and much-feared, AI revolution won't be an overnight robot takeover.
Johnson approved the plan, but the use of a new technology had Disney attorneys hammering out details over how it could be deployed, what security precautions would protect the data and a host of other concerns. They also worried that the studio ultimately couldn't claim ownership over every element of the film if AI generated parts of it, people involved in the negotiations said.
Disney and Metaphysic spent 18 months negotiating on and off over the terms of the contract and work on the digital double. But none of the footage will be in the final film when it's released next summer.
A deepfake Dwayne Johnson is just one part of a broader technological earthquake hitting Hollywood. Studios are scrambling to figure out simultaneously how to use AI in the filmmaking process and how to protect themselves against it. While executives see a future where the technology shaves tens of millions of dollars off a movie's budget, they are grappling with a present filled with legal uncertainty, fan backlash and a wariness toward embracing tools that some in Silicon Valley view as their next-century replacement.
The Academy of Motion Picture Arts and Sciences is surveying members on how they use the technology. Studio chiefs are shutting down efforts to experiment for fear of angering show-business unions on the eve of another contract negotiation. And no studio stands to gain or lose more in the outcome than Disney—the home of Donald Duck, Belle, Buzz Lightyear and Stitch, among countless others—which has churned out some of the most valuable, and protected, creative works in the world over the past century.
Interviews with more than 20 current and former employees and partners present an entertainment giant torn between the inevitability of AI's advance and concerns about how to use it. Progress has at times been slowed by bureaucracy and hand-wringing over the company's social contract with its fans, not to mention its legal contract with unions representing actors, writers and other creative partners.
The company's early steps have included discussions about adding features within its Disney+ streaming service to enable subscribers to create their own clips of Disney shows and allowing gamers to interact with an AI-generated Darth Vader in Fortnite. Disney took a $1.5 billion stake in Fortnite's owner, Epic Games, last year. Some of its efforts to use AI in movies have gone nowhere.
Meanwhile, Disney's legal teams remain fiercely protective of the studio's characters and wary of any moves internally or by third parties that could harm its brand, leading the company to sue one AI provider in June. And Disney employees who want to feed corporate information into generative AI tools for company business must first seek approval from an AI committee. (Over the past several months, the company has gotten response times to such requests down to 48 hours, according to a person familiar with the situation.)
For Disney, protecting its characters and stories while also embracing new AI technology is key. 'We have been around for 100 years and we intend to be around for the next 100 years," said the company's legal chief, Horacio Gutierrez, in an interview. 'AI will be transformative, but it doesn't need to be lawless."
It's just one of the challenges facing Chief Executive Bob Iger, who is expected to name a successor in early 2026 after nearly 20 years at the helm. He is under tremendous pressure to reduce costs amid declines in movie theater attendance and increases in people canceling their cable subscriptions. He has recently had to fend off criticism that his company has had creative challenges, and has focused too much on recycling old properties.
Disney Chief Executive Bob Iger in July.
Concerns about bad publicity were a big reason that Disney scrapped a plan to use AI in 'Tron: Ares"—a movie set for release in October about an AI-generated soldier entering the real world.
Since the movie is about artificial intelligence, executives pitched the idea of actually incorporating AI into one of the characters in the sequel to the 1980s hit movie 'Tron" as a buzzy marketing strategy, according to people familiar with the matter. A writer would provide context on the animated character—a sidekick to Jeff Bridges' lead role named Bit—to a generative AI program. Then on screen, the AI program, voiced by an actor, would respond to questions as Bit as cameras rolled.
But with negotiations with unions representing writers and actors over contracts happening at the same time, Disney dismissed the idea, and executives internally were told that the company couldn't risk the bad publicity, the people said.
Iger and Gutierrez have met with White House officials in recent months to discuss worries about AI models infringing on the company's intellectual property and using the studio's characters in inappropriate ways, according to people familiar with the discussions.
Disney is up against companies like OpenAI and Google, which argue that having access to copyrighted materials as they train their models is crucial to compete in the AI race with China, which carries national security implications.
In June, Disney and Comcast's Universal took their most drastic step yet toward protecting their creative works, suing AI provider Midjourney for allegedly making copies of their copyrighted properties.
'Midjourney is the quintessential copyright free-rider and a bottomless pit of plagiarism," the companies said in its complaint filed in June. Midjourney hasn't responded to the suit.
The lawsuit is seen by Disney's competitors as the strongest effort yet to establish a legal framework for AI issues. Some former employees and business partners who go toe-to-toe with Disney over use of its characters jokingly refer to the company as 'the largest law firm in California."
Disney said it is balancing the desire by its executives to move quickly on AI with the need to protect its characters. 'Our job is to enable our creators to use the best AI tools available without compromising the company long term," Gutierrez said.
AI is a new battleground for an entertainment company that spent the past decade weathering competition from Silicon Valley in the form of streaming rivals at Apple and Amazon, and the gravitational pull of eyeballs toward Google's YouTube.
Google, OpenAI and others now offer video, photo and sound editing tools that let fans be their own producers, with the ability to manipulate characters and images at will. That's tough for a company that controls its IP as tightly as Disney, longtime executives said. In the suit filed against Midjourney, Disney and Universal included AI-generated images of some of their most popular characters, including the Minions and Darth Vader.
Executives are keenly aware of how fast AI is advancing.
At Disney's annual management retreat in Orlando earlier this year, Rob Bredow, a senior vice president at Disney's Lucasfilm, gave a presentation showing the rapid advances of generative AI tools, some of which can generate images and scenes that, to the casual eye, appear as good as professional productions.
Bredow showed clips an artist created depicting a droid landing on a planet and the creatures it saw. Bredow explained the artist had first created the clips in the fall, and then again just a few months later, and marveled at the leap in quality of the videos, according to a person at the meeting.
Historically, Disney has been reluctant to allow its characters to mingle on consumer goods it manufactures, let alone mix and match in ways that AI tools encourage. When princesses like Cinderella and Sleeping Beauty are featured on the same product, like a lunchbox or poster, designers must ensure their gazes are fixed in different directions, so that the characters are all plausibly living in their own 'universe."
It took years for the company to allow Disney's characters from different universes—such as C-3PO and Ariel from 'The Little Mermaid"—to interact with each other in videogames.
Disney's concerns about control of its characters and stories have been a point of debate in its recent work with Epic Games, the company behind Fortnite, with tens of millions of monthly active users. Disney sees gaming as an important avenue for building future fandom.
Fortnite collapses franchises into one storytelling universe—where Batman can coexist with Lara Croft and Frankenstein's Bride. Disney is planning its own world (internally code-named 'Bulldog") connected to Fortnite where gamers can interact with characters including Marvel superheroes and 'Avatar" creatures, people familiar with the plans said.
Some Epic executives have complained about the slow pace of the decision-making at Disney, with signoffs needed from so many different divisions, said people familiar with the situation. And an experiment to allow gamers to interact with an AI-generated Darth Vader was fraught. Within minutes of launching the AI bot, gamers had figured out a way to make it curse in James Earl Jones's signature baritone. Epic fixed the workaround within 30 minutes. Ten million players spoke with Vader at least once, according to Epic.
The joint venture is being overseen by Josh D'Amaro, head of Disney's parks and resorts, who is on a shortlist of internal candidates to succeed Iger. D'Amaro has made it a point to spend time with Epic CEO Tim Sweeney, periodically visiting with him in Cary, N.C., where Epic is based, and going on hikes together, according to one of the people.
Some Disney executives have raised concerns ahead of the project's launch, anticipated for fall 2026 at the earliest, about who owns fan creations based on Disney characters, said one of the people. For example, if a Fortnite gamer creates a Darth Vader and Spider-Man dance that goes viral on YouTube, who owns that dance?
Those ownership concerns extend to major motion pictures across Hollywood. On a typical visual-effects contract, the company creating the effects gives ownership of the material to the studio. But similarly transferring ownership of AI-generated work isn't so seamless, lawyers said.
In the absence of any legal precedent, studios fear a future in which they don't own every element of a finished film, and no studio attorney wants to be the one to unwittingly let that happen.
The stakes are high for a company with as many well-known characters as Disney, which Gutierrez said doesn't want AI firms to pay for use of its characters and then assume free rein.
'We want Darth Vader just for Disney—we are not interested in surrendering control of our characters and IP to others in exchange for a check," Gutierrez said.
In some corners, the technology is embraced as a lower-cost, more efficient tool. On Amazon's 'House of David," an animated show about the biblical figure, creator Jon Erwin has boasted of the technology's godlike assistance in creating whole sequences of certain episodes.
Lionsgate, the studio behind the John Wick franchise, last year announced a licensing deal with generative AI company Runway in exchange for a custom-built AI model it can use for production. A24, the studio behind 'Everything Everywhere All at Once," was among the first studios to use Runway's AI and has hired a former expert from Adobe to help craft its strategy.
Some in the industry are scared of the technology in a way enthusiasts criticize as naive. On some sets, visual-effects crews are warned on their first day to not even mention the term 'AI." Actors who are scanned head-to-toe for digital double creation can have a representative from the Screen Actors Guild with them during the process.
Hovering over any major studio decision regarding the technology: contract talks with the Screen Actors Guild set to resume next year. Executives are reluctant to make any announcement that might anger the union or be reversed under the new contract's terms.
The 2024 movie 'Here," a Sony release that told a story spanning decades, used generative AI to de-age stars Tom Hanks and Robin Wright—and the software allowed them to see the footage of their younger selves instantaneously. When it came time to promote the film, producers grew concerned about potential pushback to having an A-list name like Hanks speak about the AI technology used in its making, a person involved in the film said.
Hanks joked about those concerns during an appearance on 'The Late Show with Stephen Colbert," asking the house band to play a foreboding sequence of notes any time he used the term AI.
'Everybody gets scared," he said.
Disney's own history speaks to how studios have navigated technological crossroads before. When Disney hired Pixar to produce a handful of graphic images for its 1989 hit 'The Little Mermaid," executives kept the incorporation a secret, fearing backlash from fans if they learned that not every frame of the animated film had been hand-drawn.
Such knowledge, executives feared, might 'take away the magic."
Write to Jessica Toonkel at [e-mail] and Erich Schwartzel at [e-mail]