Runway builds custom AI video models for Lionsgate

Hollywood studios spent the last year fighting over copyright and protesting generative tools, but one major player just decided to feed its entire archive to an AI company.

Video generation startup Runway announced a first-of-its-kind partnership with Lionsgate. The two companies are building and training a custom AI model powered exclusively by Lionsgate’s proprietary catalog of over 20,000 film and television titles.

The customized model is fundamentally designed for Lionsgate Studios, its filmmakers, directors, and creative talent. It will generate cinematic video that can be dialed in and adjusted using Runway’s existing suite of controllable editing tools.

Cutting costs across pre-production and post

Lionsgate Vice Chair Michael Burns framed the deal as a way to unlock capital-efficient content creation. In plain English, that means finding faster, cheaper ways to get movies and shows made.

Burns noted that several Lionsgate filmmakers are already exploring the tool for both pre-production storyboarding and post-production workflows. The studio positions the tech as an augmentation layer rather than a wholesale replacement for human teams.

Runway co-founder and CEO Cristóbal Valenzuela highlighted that the partnership gives storytellers better tools to bring ideas to life. He argued that the history of art has always been the history of technology.

Runway also dropped an interesting hint for smaller creators. While currently working with enterprise studios like Lionsgate, Runway noted it is exploring ways to license these custom models as templates. That could eventually allow individual creators to train proprietary models on their own personal archives.

Why it matters

We think this partnership marks a major shift in how Hollywood deals with generative video.

Instead of fighting endless copyright lawsuits over scrapers ingesting protected movies, Lionsgate took the licensed, walled-garden route. Training a model strictly on 20,000 owned titles gives the studio legal clarity while letting Runway produce cinematic video tuned specifically to Lionsgate's production quality and intellectual property.

For everyday creators and indie filmmakers, the immediate takeaway is Runway’s nod toward private model templates. If Runway rolls out custom fine-tuning tools to individual artists, creators could soon train models on their own past films or animation styles without needing a server farm.

However, there is an obvious tension here. When an executive highlights capital efficiency in film production, it usually points toward leaner crews. Even if Lionsgate calls this tool an augmentation for talent, roles in concept art, storyboarding, and junior visual effects will inevitably face pressure as other legacy studios inevitably follow suit.