Katzenberg launches AI video startup with ex-OpenAI Sora leader

Katzenberg Returns to the AI Arena with a New Video Venture

Hollywood mogul Jeffrey Katzenberg — co-founder of DreamWorks Animation and former head of Disney Studios — is joining forces with Tim Brooks, the former leader of OpenAI's Sora video generation project, to launch a new artificial intelligence video startup. The pairing unites one of entertainment's most recognizable dealmakers with one of the researchers most closely associated with the recent explosion in AI-generated video technology. Katzenberg has already made a high-profile, if turbulent, mark on the AI-adjacent media landscape through his involvement with Quibi, the short-form streaming platform that collapsed spectacularly in 2020 after burning through nearly $2 billion. This new venture signals a deliberate return — and a bet that generative AI video has reached a commercial inflection point.

Brooks departed OpenAI after leading the team behind Sora, a text-to-video model that drew widespread attention when it was demonstrated publicly in early 2024. Sora was capable of producing remarkably fluid, visually coherent video clips from text prompts, and its release was treated as a watershed moment for AI-generated media. Brooks was among the more prominent technical minds at OpenAI, and his exit reflects a broader pattern of senior researchers and product leaders leaving the company to pursue independent ventures — a trend that has seen OpenAI alumni fan out across the startup ecosystem, often raising substantial capital quickly given their pedigree.

Why This Partnership Is More Than a Vanity Project

The combination of Katzenberg's Hollywood network and Brooks's deep technical background is notable precisely because AI video has struggled to bridge the gap between impressive demos and practical production-grade tools. Katzenberg spent decades navigating the economics of animation and film production, overseeing franchises like Shrek and Madagascar, and later co-founding WndrCo, a holding and investment company. His involvement suggests the new startup is aiming squarely at professional and commercial media production rather than the consumer novelty market.

Brooks, for his part, brings firsthand experience scaling one of the most complex generative models ever deployed. Sora required enormous compute resources and presented novel challenges around temporal consistency — keeping objects, lighting, and physics coherent across video frames — problems that competing systems from Runway, Stability AI, and Google's Lumiere project have all grappled with. If Brooks applies lessons from Sora's development to a well-funded startup environment without the constraints of a larger corporation's priorities, the resulting product could move faster and be more tightly targeted to industry needs.

The startup enters a market that is simultaneously crowded and still wide open. Runway ML has become the most prominent independent AI video company, raising over $230 million and cultivating a loyal base of professional video editors and filmmakers. Pika Labs emerged as a fast-moving challenger. Meanwhile, major technology companies — Google with Veo, Meta, and Adobe with its Firefly video tools — are embedding video generation directly into existing creative software suites. For a new entrant to carve out space, it will likely need either a decisive technical advantage, a distribution edge through entertainment industry relationships, or both.

The Stakes for AI-Generated Media

The timing of the Katzenberg-Brooks venture coincides with growing urgency inside Hollywood about what AI video actually means for the production workforce and the economics of content creation. The 2023 writers' and actors' strikes placed AI use at the center of labor negotiations, with studios and streaming platforms seeking latitude to use AI tools for everything from scriptwriting assistance to digital likeness replication. Unions won some protections, but the agreements left considerable ambiguity about how AI-generated imagery and video would be governed in future productions.

A startup with Katzenberg's access to studio relationships and talent could navigate that landscape in ways that a purely technical team might not. At the same time, that proximity to Hollywood could make the company a lightning rod for criticism from creative labor advocates who are already wary of AI's displacement effects. The challenge will be positioning tools as augmentations for human creators rather than replacements — a framing that every major AI company has adopted but that remains deeply contested in practice.

From a purely financial standpoint, the venture arrives at a moment when investor appetite for AI infrastructure and applications remains intense despite broader venture capital caution. OpenAI alumni carry particular fundraising credibility, and Katzenberg's name opens doors to both entertainment-sector strategic investors and traditional Silicon Valley venture funds. The combination makes a substantial seed or Series A raise appear straightforward, though the longer-term question is whether the company can build recurring revenue from professional customers before the largest technology companies commoditize the underlying video generation capability.

What Comes Next

The new company does not yet have a public name or announced product, and the full scope of its technical ambitions remains unclear. Key open questions include whether Brooks is bringing any additional OpenAI talent with him, what specific use cases the startup will prioritize first, and how it plans to handle the thorny issues of training data rights that have entangled competitors in litigation. Shutterstock, Getty Images, and various artist groups have all pursued legal action against AI image and video companies over the use of copyrighted material in training datasets, and any new entrant will face immediate scrutiny on that front.

Katzenberg's history suggests he thinks in terms of large, transformative bets rather than incremental product iterations. His willingness to stake his reputation again — after the Quibi debacle subjected him to years of pointed criticism — indicates genuine conviction that AI video is approaching a moment of commercial and cultural significance that rivals earlier platform shifts in the media industry. Whether Brooks's technical depth and Katzenberg's industry leverage add up to a durable business will depend heavily on how quickly the broader AI video market matures and who controls the distribution channels when it does.