Google DeepMind and A24 have announced a multi-project research partnership that will put AI researchers directly into the filmmaking workflow, pairing one of the world’s most influential AI labs with the studio behind a generation of internet-native prestige films.
The companies announced the partnership on June 22, saying the work will focus on new creative workflows and techniques shaped with A24 filmmakers over time. Google has also made an investment in A24. The companies did not disclose the size of that investment in the announcement, but the Wall Street Journal reported it is around $75 million and marks the first time Google has taken an equity stake in a film studio.
The news matters because it moves the AI-and-entertainment debate away from abstract claims about replacing artists and toward a more specific test: whether a major AI lab can build tools inside a real production culture without turning a studio’s catalog, directors, and audience trust into raw material for generic generation.
What Google and A24 actually announced
Google’s official description is deliberately broad. The partnership is framed as a research and development collaboration between A24 and Google DeepMind that will span multiple projects. A24 filmmakers are expected to test ideas, give feedback, and help shape the technology as it is built, rather than receiving a finished product after the fact.
That distinction is important. The announcement does not name a commercial product, launch a consumer AI video app, or describe a plan to generate complete films from prompts. It describes a working relationship around production techniques, creative workflows, and future entertainment tools whose goals will evolve over time.
Independent reporting adds several useful constraints. The Verge reported that the deal is non-exclusive, that Google will not receive access to A24’s film and television library, and that no specific public projects have been disclosed. That library-access point will matter to artists and rights holders watching the deal closely, because it separates the partnership from the broader fight over training data and studio archives.
The Wall Street Journal also reported that A24 has a roughly 20-person innovation team, A24 Labs, and is already working on AI-generated storyboards. If that is where the early work lands, the first practical use case may look less like automated moviemaking and more like faster previsualization: turning early concepts into visual references that directors, production designers, cinematographers, and producers can debate before a shoot gets expensive.
Why A24 is a meaningful partner
A24 is not a conventional studio partner for a technology company. Its value is tied heavily to taste, filmmaker relationships, audience identity, and a brand that younger moviegoers often read as artist-first. That makes the partnership more sensitive than a back-office AI deal at a large media conglomerate.
It also makes the partnership useful for Google. Building tools with A24 gives DeepMind access to a messy, high-judgment creative environment where technical demos are not enough. A storyboard tool, for example, has to respect tone, camera language, pacing, continuity, and the director’s visual vocabulary. A distribution or marketing tool has to understand audiences without flattening every film into a performance-optimized content object.
A24’s risk is different. Its audience is unusually attuned to authenticity, labor, and corporate influence. A partnership with Google and DeepMind could help the studio develop tools that fit how filmmakers actually work. It could also create skepticism if the results look like a studio borrowing the aesthetics of independent film while letting a tech partner define parts of the creative process.
The hard part is creative control, not just generation quality
Most AI video arguments still get stuck on whether generated clips look good. For studios, the harder question is whether AI systems can be useful without blurring authorship, rights, labor roles, and control over a film’s style.
Filmmaking is full of tools that already shape creative decisions: editing software, visual effects pipelines, digital color grading, motion capture, virtual production, previs, and marketing analytics. AI will likely enter through those same production layers before it replaces them. That is why storyboards, concept development, localization, asset search, rough edits, audience testing, and distribution planning are more plausible first targets than an end-to-end AI movie generator.
The partnership also lands against a tense industry backdrop. Writers, actors, directors, studios, and AI companies are still negotiating the boundaries of consent, likeness, training data, residual value, and credit. A deal that keeps A24’s library outside Google’s reach may calm one major concern, but it does not answer every question. Filmmakers will still want to know what data new tools learn from, how project materials are stored, whether outputs are traceable, and how much control artists have over model behavior.
What to watch next
The first real signal will be whether A24 and DeepMind describe a concrete production tool rather than another broad research promise. A storyboard generator would be a logical starting point because it can speed early planning while leaving final creative decisions with human teams. But even that use case raises practical questions: can directors tune style without copying living artists, can production teams track AI-assisted frames, and can unions see where the tool changes work on set?
The second signal will be who uses the tools. If A24 quietly tests them on internal development work, the partnership may stay mostly experimental. If named directors, producers, editors, or production designers begin using the system on real films, the deal becomes a more serious benchmark for creative AI adoption.
For Google, the partnership is a chance to make DeepMind’s creative AI work look less like a lab demo and more like infrastructure for professional media. For A24, it is a bet that artists can shape the tools early enough to avoid being shaped by them later. The outcome will depend less on whether AI can make striking images and more on whether filmmakers trust the system when the work is still fragile, expensive, and unmistakably theirs.
