AI's Hollywood Push Runs Into Limits

Generative-AI companies and Silicon Valley giants are increasingly trying to embed themselves in Hollywood workflows — pitching text-to-image and text-to-video tech as production shortcuts even while the tools remain limited. As image models improved, studios started talking to Runway, OpenAI, Google, and Meta about partnerships; OpenAI’s high-profile move is a planned feature-length, AI-driven project called Critterz, while Lionsgate struck a deal with Runway and Amazon invested in Showrunner.

The pitch is simple: AI can lower costs and democratize production at a time when studios are trimming budgets and greenlights are harder to get. Some established filmmakers are signing on to AI-forward projects, and streaming services like Netflix are openly experimenting with generative tools.

But the technology has glaring limits. Most models produce only seconds of inconsistent footage, offer little fine control, and struggle when trained on a single studio’s catalog (reportedly a problem in the Lionsgate–Runway effort). Legal fights over copyright, lawsuits from studios, and labor concerns about job losses are also major obstacles. And cost arguments are mixed — OpenAI’s Critterz is rumored to cost $30M, well above some low-budget projects made with traditional open tools like Blender.

The result: eager business development and high-profile demos, but uncertain creative and legal outcomes. For gen‑AI to truly break into Hollywood it will need technical reliability, clearer rights frameworks, and answers to the industry’s labor concerns.