How Will Hollywood Contend With AI Without Destroying Itself?

Share this articleWhile finishing Mercy, his upcoming film starring Chris Pratt, Timur Bekmambetov decided to make a new movie almost entirely generated by artificial intelligence. The 64-year-old has directed more than a dozen films, including the Angelina Jolie action thriller Wanted and a remake of Ben-Hur, and has always been an early adopter of new techniques. He’s produced a few films, such as 2018’s Searching, that take place almost entirely on smartphones and computer screens.Generative AI video technology is not yet good enough for anyone to make an entire feature-length movie. But Bekmambetov felt compelled to try. He adapted The Man With a Shattered World, a book about a young soldier who sustains a traumatic brain injury in battle and the efforts to recover his memories. The project, Bekmambetov surmised, could make a virtue of AI’s shortcomings—its glitches and hallucinations could be presented as reflections of the soldier’s brain damage.What would have been a $150 million movie employing hundreds of people cost a fraction of that, featuring just one actor and a skeleton crew. While the result isn’t ready to be released—Bekmambetov says this is just a proof of concept—he’s convinced AI is going to upend the way people make movies. He’s even developing software that will allow anyone working on a film project to integrate AI into their work, creating a digital assistant to help each department. “I am 100% sure it’s a revolution. Not in cinema—in every industry,” he says from his home base of Cyprus. “We cannot stop it.”Not everyone is as willing to embrace this vision of the future. Bekmambetov recently got a call from one of the producers of Mercy, a thriller about a detective who has 90 minutes to convince a judge that he didn’t murder his wife. The judge, Maddox, is an artificially intelligent digital being whose avatar presents as Rebecca Ferguson. In spite of the plot, Amazon completely forbade the use of AI on the film, which the company will release in January. “He called me and said the actors are not happy that you are trying to create an AI service,” Bekmambetov says.Rebecca Ferguson plays an AI avatar in Amazon’s Mercy. Photographer: AlamyThe rise of artificial intelligence has divided Hollywood, a contracting industry torn between its desire to experiment with new technology and its fear of its implications. While development executives use ChatGPT to analyze scripts and marketers use it to assist with creative campaigns, many of those same people are worried their companies will use AI to eliminate their jobs. A growing chorus of filmmakers has come out against generative AI, including director Guillermo Del Toro, who said in October that he’d rather die than use the technology in his films.This tension could well boil over in the coming year. December saw the industry's biggest move yet, with Walt Disney Co. entering a licensing agreement for OpenAI’s Sora and investing $1 billion in the company. Until now, studios have been eager to tout the potential benefits of AI to investors, but afraid to divulge their biggest experiments, lest they antagonize talent and alienate labor unions. Writers and actors went on strike for months in 2023, animated in no small part by concerns over AI. Those contracts expire again in 2026, and Hollywood is bracing for another potential labor stoppage. “Most of the studios have been too timid,” says Amit Jain, CEO of Luma AI, a leading video generator. “They’re scared to talk to their filmmakers or to bring AI to them.” They have also been wary of AI’s potential impact on their most valuable assets, their film and TV catalogs.Hollywood has survived a century of technological advances by integrating sound, color and visual effects, but the age of computing has undermined the foundation of the entertainment business unlike any innovation before. Despite the industry’s attempts to fight new players through litigation and regulation, the internet eliminated studios’ stranglehold on video distribution. Napster ushered in an era of file-sharing sites that allowed movie fans to access blockbusters for free, before giving way to Netflix, YouTube and other legal streaming platforms, which offered bottomless entertainment far cheaper than cable TV.When OpenAI released the latest edition of its video app, Sora 2, in October, it augured a world in which AI will eliminate the industry’s advantage in creating that entertainment as well. The smartphone has already put a camera in the pocket of every aspiring Spielberg. New AI models, trained on vast amounts of video, including hit movies and TV shows, could enable those young filmmakers to produce work that resembles Hollywood content for much less money and in much less time.The industry initially erupted in anger in the fall, convinced that OpenAI had violated its rights by allowing anyone to produce videos featuring famous people and copyrighted characters unless the stars or studios opted out. While no lawsuits were filed, the uproar pressured the company to revise its policy by requiring an opt-in from celebrities and IP holders.The growing sense that AI video would thunder ahead pushed studios to begin publicizing small ways they’re implementing the technology. Netflix said it used AI to de-age characters in Happy Gilmore 2 (see above). Amazon instituted a team to find ways AI might create efficiencies in areas like animation and dubbing. Disney teased it would soon allow fans to use the technology to create videos of its princesses and superheroes—its landmark OpenAI deal followed a month later.Having been slow to embrace many digital technologies, entertainment leaders say they are trying to learn from past mistakes. They may have no choice. OpenAI, Google and Meta have the resources to fight any lawsuit, while governments will take too long to act.“There is a recognition that you can’t put the genie back in the bottle,” says Aaron Moss, a partner at Mitchell, Silberberg & Knupp and author of the blog Copyright Lately. “These lawsuits take a long time to resolve.” The most effective tactic is to impose limits through negotiations and find ways to utilize the technology. AI could produce useful tools that make films look even better. It could also reduce the cost of production, which has ballooned in recent years.Actors, writers and directors are worried AI will devalue human work and jeopardize their livelihoods, and they are sounding the alarm at the same time studios have started to talk about their AI moves. Knives Out director Rian Johnson told the Hollywood Reporter that the technology is “making everything worse in every single way.” At a recent film festival in Marrakech, actress Jenna Ortega and filmmakers Bong Joon Ho and Celine Song all spoke out against the use of AI, with Ho joking that he would organize a military squad to destroy the technology.While the Hollywood workers at immediate risk are those in less glamorous jobs like animation, visual effects and makeup, the implementation of AI could be stickiest when it comes to onscreen talent. A studio owns the rights to a movie, but it doesn’t own the rights to Tom Hanks’ face or Lady Gaga’s voice. And while there is an established legal framework for copyright, the rights to publicity, name, image and likeness are much fuzzier.Talent representatives at major agencies, management companies and law firms are studying the potential implications for their clients. Kevin Yorn, a prominent entertainment lawyer, demands clauses in every client’s contract to govern the use of digital replicas and synthetic likenesses, ensuring that a studio can’t keep using a digitized version of them for future projects without permission. He’s even filed with the copyright office to secure the voice of one of his clients, an uncertain legal maneuver.“AI means someone can re-create your voice, your face, your movement, your cadence, your entire persona from past performances or scrape media without you,” says Yorn, whose clients include Scarlett Johansson. The star, who played an AI love interest in the movie Her, claimed OpenAI mimicked her voice in that film for its voice assistant after she declined to participate in the project. As Yorn sees it, AI poses an existential threat that “collapses the boundary between a person and a file.”Despite his reservations, Yorn is urging his clients to experiment. Matthew McConaughey, another client, recently teamed up with the startup Eleven Labs so it can clone his voice. McConaughey also invested an undisclosed amount of money in the company, which could help big stars make even more money than they do today by introducing a new revenue stream. The actor has already served as a spokesperson for Uber Eats and Salesforce and a voice actor in Sing, but what if his voice could be licensed by any advertising company or studio? “AI isn’t a theoretical thing,” Yorn says. “It’s already embedded in how Hollywood is operating now.”Lede: In Netflix’s Happy Gilmore 2, Adam Sandler was de-aged using artificial intelligence. Animation: 731; photos: Netflix