Google DeepMind and A24 announced a multi-project research partnership today that embeds the studio's filmmakers inside DeepMind's generative AI development, and Google has taken an equity stake in A24 as part of the deal. The financial terms were not disclosed. It is the first named creative-industry partnership of this depth for DeepMind, and the first time A24 — the studio behind Everything Everywhere All at Once, Uncut Gems, and The Brutalist — has tied itself to a single AI research lab.
The structure matters more than the announcement. Rather than licensing footage or renting API access, A24 filmmakers will sit alongside DeepMind researchers to shape the tools as they are built. DeepMind said the goal is to help artists develop new workflows while giving its researchers direct feedback from working directors and craftspeople.
“The collaboration pairs a world-leading research lab with the industry's most filmmaker-forward studio to help artists develop new workflows and techniques.”— Google DeepMind, official announcement
That model addresses the loudest complaint from working filmmakers about generative video: the tools are built by engineers who do not shoot for a living, and ship with defaults that assume a stock-footage use case rather than a directed one. Runway, Luma, and OpenAI's Sora have all leaned on outside filmmaker advisory boards for feedback, but a deep bilateral partnership with a single major studio is a different commitment on both sides.
Key facts
- 01Google DeepMind and A24 announced a multi-project research and development partnership focused on filmmaking tools.
- 02Google has taken an equity stake in A24 as part of the deal; the size was not disclosed.
- 03A24 filmmakers will work side-by-side with DeepMind researchers to test and iterate on generative AI tools.
- 04The partnership has no fixed deliverables — goals and outputs will evolve over time.
For A24, the calculus is defensive and offensive at once. The studio has built a brand on filmmaker autonomy, and its directors are among the loudest voices arguing that generative tools need to bend to a creative vision rather than dictate one. Getting inside the lab, rather than reacting to whatever ships, is the way to make that argument stick in the code.
For DeepMind, the deal slots in alongside its existing generative-video work on Veo and its image work on Imagen, both of which have been folded into Google's consumer products through Gemini. Neither the blog post nor the announcement named specific models the partnership will target, and DeepMind explicitly said the technical outputs will evolve over time — meaning this is a research relationship, not a product launch.
The vagueness is the point. DeepMind framed the deal as the beginning of a collaborative journey rooted in research and shared curiosity, with no fixed deliverables and no committed release schedule. That is unusual for a Google announcement, which typically leads with a shipping product, and it signals that both sides expect the useful output to be workflows and internal tools rather than a public model release.
Google's investment in A24 is the harder-nosed piece. A24 has raised outside capital before — Stripes Group led a round in 2022 that valued the studio at roughly $2.5 billion — but a strategic check from Google puts one of the largest AI companies on the cap table of one of the most prestige-heavy independent studios in Hollywood. It also creates a template competitors will study. OpenAI, xAI, and Meta all have generative-video ambitions and no comparable studio relationship.
The deal lands during an unresolved fight in Hollywood over what generative AI can and cannot do with union labor, and over whether models were trained on copyrighted footage without permission. A24 is a signatory to the same guild agreements as the major studios, and any tools that emerge from the partnership will have to work inside those constraints. Neither DeepMind nor A24 addressed training data or guild scope in the announcement, and that silence is likely to draw questions from SAG-AFTRA and the DGA before it draws answers.
The other open question is what the tools actually do. Generative video has advanced quickly on short clips and stylized shots, and has advanced less on the things narrative filmmakers care about most: consistent characters across scenes, precise camera control, and integration with the existing edit-and-VFX pipeline. If the partnership produces meaningful progress on any of those three, it will move the market. If it produces another clip generator, it will not.
Google buying its way into A24 is a bet that the next round of generative-video competition will not be won on benchmark scores but on which lab is trusted by the people paying to make actual films. That is a very different playing field from the one OpenAI and Runway have been fighting on, and it is one where DeepMind's parent company has an unusual asset to spend: the cash to take an equity stake in a studio it wants as a design partner. Expect the other frontier labs to go shopping.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
