← All posts / Industry

Google Is Courting Hollywood to License Film IP for AI — and the Deals Are Stalling

Google has approached Disney, Universal and Warner Bros. Discovery about licensing characters and film libraries for AI models, but union rules, talent-rights sensitivities and a demand that participants waive the right to sue have stalled every deal.

Google Is Courting Hollywood to License Film IP for AI — and the Deals Are Stalling

A decade after Hollywood treated Google as a pariah over pirated clips on YouTube, the search giant is back at the studio gates — this time with a checkbook instead of a takedown notice. According to a Los Angeles Times investigation published on August 31, 2026, Google executives have approached representatives at Disney, Universal, Warner Bros. Discovery and other studios about licensing intellectual property for use in AI models. Three separate tracks of negotiation are underway: using Google’s AI tools inside film and TV production, licensing iconic studio characters for AI-generated outputs, and building YouTube-based systems to detect and monetize synthetic content and unauthorized celebrity likenesses.

So far, not a single major-studio agreement has been reached. And the reasons why say a great deal about where the AI industry’s thorniest problems now live: not in model architecture, but in contracts, labor agreements and the right to sue.

What Google is actually asking for

The most concrete revelation in the report is pricing. One person familiar with these deals estimated that an AI company licensing character IP from a studio could pay around $40 million per character on average — meaning a package covering 100 characters would run into the multiple billions of dollars. That figure reframes studio libraries as potentially among the most valuable untapped training and generation assets in the entertainment economy.

The character-licensing track could open a new revenue stream for studios, while Google would share in advertising revenue from AI-generated content distributed on YouTube. The third track is perhaps the most strategically interesting: YouTube has approached talent agencies and studios about deploying its likeness detection technology to flag AI-generated content that infringes copyright or alters celebrity faces — with a mechanism to let rights holders monetize synthetic videos rather than remove them. Justin Connolly, YouTube’s global head of media and sports and a former Disney executive, is reportedly involved in those discussions.

But the terms attached to that detection program are where the talks hit a wall. Sources told the Times that some companies were asked to waive their right to sue Google as a condition of participating in the likeness-detection effort. For an industry that is actively litigating against AI companies — and where Disney sent Google a cease-and-desist letter last year accusing it of training AI systems on Disney’s copyrighted material without authorization — surrendering legal leverage is a red line few studios will cross blindly.

Why now: the Sora opening

Google’s timing is not accidental. OpenAI shut down its text-to-video app Sora this year to focus resources elsewhere, and with it collapsed a landmark licensing deal Disney had announced with OpenAI that would have brought AI-generated versions of characters like Mickey Mouse to the platform. That retreat left a vacuum in legitimate, studio-sanctioned generative video — and Google is positioning to fill it.

The company has spent more than a year building credibility in Hollywood. In June it invested $75 million in A24, the studio behind Everything Everywhere All at Once and Moonlight, as part of a strategic partnership to collaborate on AI tools. It also maintains a partnership with director Darren Aronofsky’s venture Primordial Soup. “They recognize that A24, while not the biggest studio, gives them a type of credibility they’re not getting yet in Hollywood,” said Rob Rosenberg, a partner at Moses & Singer and former Showtime general counsel. “They’re in this image rehabilitation, this path of trying to gain not just acceptance but appreciation.”

An industry insider quoted by the Times put the technical stakes bluntly: “In order for Google’s tools and technology to be truly usable for large-scale film and TV productions, let alone cinematic exhibition, it is essential that their models be fine-tuned with highly dynamic content and metadata that only Hollywood studios have.”

The union constraint

The quiet blocker in all of this is labor. Studios operate under contracts with AI protections for writers, directors and actors — and under its agreement with the performers’ union, any new AI licensing deal triggers a notification requirement to SAG-AFTRA. Duncan Crabtree-Ireland, the guild’s national executive director and chief negotiator, confirmed that no major studio has notified SAG-AFTRA of entering any new AI licensing deal.

“The companies largely are moving carefully in this area,” Crabtree-Ireland said. “They rightly recognize that there are a lot of potential pitfalls, not only relating to creative talent and their rights and performances or other types of creation.”

Actors and writers have complained that their work was used to train AI models without permission or compensation, and visual-effects workers fear outright replacement. Studios that move too fast risk a labor revolt; studios that move too slowly risk being left behind by technology that makes production faster and cheaper.

History rhymes — at $40 million a character

Veterans of the last Google-Hollywood war are watching the rhyme closely. In 2007, Viacom sued Google for $1 billion over copyright infringement on YouTube; the case settled in 2014, by which point YouTube was too central to the media ecosystem for studios to avoid. Today, studios market their programming on YouTube, and Google’s Content ID system flags infringing videos and lets rights holders earn advertising revenue from them.

“What Google did with YouTube is it let it run completely wild until it was too big to fail and everybody had to make a deal with it at some point,” said Jonathan Miller, CEO of Integrated Media and a former News Corp. executive. The lesson studios drew: never again let the technology achieve escape velocity before the contracts are signed.

Meanwhile, the broader landscape keeps moving. Disney named AI companies including Promise and Open Art to its accelerator program in July. Netflix bought Ben Affleck’s AI film tech startup InterPositive for $587 million in March. NBCUniversal — publicly, at least — says it is not in active licensing conversations with Google, and Google, Disney and Warner Bros. Discovery all declined to comment on the talks.

“We’ve moved from the moment where there was a question, ‘Will we use AI?’ and now it’s a question of ‘How are we going to use AI?’” Rosenberg said. Google’s bet is that the answer runs through Mountain View — provided it can convince Hollywood to sign away fewer rights than it is accustomed to keeping.