Hollywoods High Stakes Bet On Generative AI

The Rise of Generative AI in Film

In just a few years, generative AI has evolved at a breathtaking pace. Text-to-image models advanced from producing smudgy, abstract visuals to creating hyper-realistic images from simple descriptions. Following this leap, text-to-video tools like Runway’s Gen series, Meta’s Make-A-Video, and Google’s Veo began to mature, catching the attention of major Hollywood studios now ready to explore gen AI's potential.

Despite significant concerns around copyright infringement and job displacement, a growing chorus of advocates insists that AI is the future of filmmaking. Many proponents frame it as a democratizing force, lowering traditional creative barriers like learning to draw, play an instrument, or write a script. And while the output from AI video generators has often been lackluster, entertainment studios are increasingly betting on the technology's long-term payoff, sometimes overlooking the potential harms.

The use of AI in film production isn't entirely new, though studios have been quiet about it. Industrial Light & Magic utilized AI to de-age actors in Martin Scorsese’s The Irishman, and Marvel Studios used machine learning to superimpose actors’ faces onto stunt doubles in Shang-Chi. This trend continued with AI-assisted visual effects in films like Thor: Love and Thunder and Ant-Man and the Wasp: Quantumania.

While the public may have been unaware, many actors and writers were not. Their concerns about AI's impact on their careers were a major factor in the 2023 dual entertainment strikes. Shortly after, Disney and Marvel's Secret Invasion controversially featured an AI-generated opening sequence. Even after the strikes concluded with some AI protections, studios continued using the technology in productions like Indiana Jones and the Dial of Destiny, Dune: Part Two, and Late Night With the Devil.

Silicon Valley's Hollywood Push

Generative AI still has glaring limitations for serious video production. Most models produce only a few seconds of footage with inconsistent visual details and offer little creative control. However, this isn't deterring Silicon Valley heavyweights and a host of AI startups from trying to embed themselves within the entertainment industry.

In recent months, major players like OpenAI, Google, and Meta have been actively meeting with film studios. These efforts have led to concrete deals. Lionsgate, for instance, signed a deal with Runway to create an in-house AI model trained on its film library. Amazon invested in Showrunner, a company self-described as the “Netflix of AI” that focuses on user-created animation via text prompts. Meanwhile, OpenAI announced plans to produce a feature film called Critterz to demonstrate that entire projects can be made with generative AI.

There has also been a surge in collaborations between established filmmakers like David Goyer, Darren Aronofsky, and James Cameron and AI companies. They are pitching AI as a solution to the industry’s struggles with ballooning budgets and a depressed box office.

With Hollywood production down since 2024, and major studios reining in spending, getting new projects greenlit has become tougher. Startups like Asteria argue that by using AI, they can slash production costs dramatically, empowering filmmakers to get projects made independently.

Hurdles and Headwinds for AI in Entertainment

Despite the hype, the reality of generative AI in Hollywood remains uncertain. We have yet to see more than concept art for Critterz, and it's unclear what a film from a studio like Asteria, filled with AI assets, would actually look like. Netflix has openly embraced AI for cost-saving purposes, but landmark deals like the one between Lionsgate and Runway have been plagued with technical problems. After initial boasts, reports surfaced that Runway’s tech couldn't perform as promised because Lionsgate's entire IP library was still an insufficient dataset.

Nevertheless, Silicon Valley remains determined. Google’s push to attach its AI tech to projects like Ancestra—a short film largely composed of what looks like machine-generated stock footage—signals a clear intent to force these partnerships into existence.

For that to happen, AI companies must first contend with major legal battles. Studios like Disney and Universal are filing lawsuits over copyright infringement. These models are only as good as their training data, and many were built on the assumption that the output would be so impressive that no one would question their unscrupulous origins. But megacorporations care deeply when their IP is stolen. More importantly, filmmakers have been sounding the alarm that early AI adoption has already led to the elimination or consolidation of creative jobs. It is difficult to see how this technology can be integrated without putting a wide array of artists out of work.

To a studio executive, a machine that produces endless concept art for pennies sounds like a dream. But for the artists who once did that work, generative AI represents an existential threat.

The AI Cost Paradox

• AI boosters often claim the technology will make projects cheaper, but OpenAI's Critterz is budgeted at a reported $30 million. This is significantly more than the $4 million spent making Flow using Blender, a free open-source program, before the film went on to win an Oscar.
• Julian Glander's Boys Go to Jupiter is another testament to the power of tools like Blender. In an interview, Glander noted that the Blender community is a prime example of technology truly democratizing art by helping people learn a craft.

Dig Deeper

• Read The Wrap’s report for a detailed look at why the Lionsgate and Runway AI collaboration is faltering. The piece highlights a key issue: even a studio's entire catalog may not be enough data to produce usable output.
• To understand the hype cycle, it's worth hearing from the founders themselves. Interviews with Asteria co-founder Bryn Mooser and Fable CEO Edward Saatchi reveal how many are selling products that still seem half-baked.
• Before seeing the AI-upscaled Wizard of Oz at the Las Vegas Sphere, consider David Ehrlich’s IndieWire piece, which argues the project is less a restoration and more a mutilation of the 1939 classic.