The A24 logo on a clean dark background with subtle digital grid lines fading in from the edges, suggesting the intersection of indie cinema and technology
feature

Google Just Invested $75 Million in A24 — and It's Not Just About the Movies

Google DeepMind's first-of-its-kind partnership with A24 signals a new era for AI in filmmaking. Here's what the deal actually means for directors, writers, and the future of independent cinema.

By BucketMovies Editorial 9 min read
#A24 #Google #AI #filmmaking #Hollywood #technology

Editorial Notes

BucketMovies Editorial covers classic cinema, repertory discoveries, and context-rich film criticism with an emphasis on source-backed reporting and careful editorial review.

On Monday, June 22, Google DeepMind wired $75 million to A24 — the studio behind Everything Everywhere All At Once, Marty Supreme, and this year’s viral horror hit Backrooms. The investment, first reported by the Wall Street Journal, is being called a “first-of-its-kind” research partnership. That phrase gets thrown around a lot in press releases. But this time, the details justify it.

DeepMind is not buying distribution rights or a seat on the board. The company is building AI tools for filmmakers, and it wants A24’s directors, editors, and writers to tell it what those tools should actually do. The idea is to develop features that support “authentic, meaningful storytelling” — Demis Hassabis’s words, not a marketing intern’s — by collaborating with artists from the start rather than dropping finished products on them after the fact.

The deal lands at a moment when AI and Hollywood are circling each other with a mix of hunger and distrust. Netflix paid $587 million for Ben Affleck’s AI filmmaking startup InterPositive earlier this year. Amazon’s MGM Studios launched an internal AI unit last year focused on television and movie production tools. What makes the A24-DeepMind deal different is the studio on the other end of the check.

Why A24, and why now

A24 occupies a strange and enviable position in Hollywood. It is an independent studio that has become a brand — the way Pixar was a brand in the 2000s or Miramax in the 1990s. People who do not read film credits know the A24 logo. The studio has managed to make “indie” a commercial identity: Everything Everywhere All At Once won seven Oscars and grossed over $140 million worldwide. Marty Supreme, starring Timothée Chalamet, became one of the most talked-about films of 2025 before anyone had seen a single frame — partly because A24 put Chalamet on top of the Las Vegas Sphere as a marketing stunt.

The studio’s track record since its 2012 founding has been remarkably consistent. Moonlight won Best Picture. Uncut Gems turned Adam Sandler into an awards-season argument. Hereditary and Midsommar redefined what horror could look and feel like. Lady Bird made $79 million on a $10 million budget. The Zone of Interest won Best International Feature and Best Sound. In an era when most independent studios have been acquired, folded, or reduced to brand labels within larger conglomerates, A24 has stayed independent and kept growing. This year’s Backrooms — a YouTube-creepypasta adaptation that had no business being a theatrical hit — grossed over $100 million domestically.

That cultural cachet is the real asset Google is investing in. DeepMind could partner with any studio. It could build tools in a lab and license them. But if the first major AI-for-film partnership had been with, say, a legacy studio known for franchise sequels, the industry’s response would have been an eye roll and a panic attack. Partnering with A24 — the studio that directors like Ari Aster, the Daniels, and the Safdie brothers chose to work with — sends a different signal. It says these tools are being designed for artists, not for cost-cutting.

Hassabis framed it this way in the announcement: “We believe the best way to develop tools that empower artists is to work directly with them. By collaborating with filmmakers and industry leaders like A24 from the beginning, we can build new AI features to support artists in authentic, meaningful storytelling that helps enable their creative vision.”

Translation: we are not building a “make movie” button. We are building tools that a cinematographer or editor might actually want to use — and we want the people who made The Lighthouse and Hereditary to tell us if we are getting it right.

The technology question nobody is answering yet

What the tools actually do remains vague. DeepMind and A24 have not released a product roadmap. The press release mentions “AI features” without specifying whether those are pre-visualization tools, automated rotoscoping, dialogue generation, scene composition assistants, or something entirely new. That openness could be strategic — announcing specific tools invites specific criticism — but it also means the entire industry is watching a blank screen and filling it with their own anxieties.

The range of possibilities is enormous. On the benign end: tools that let a director sketch a shot and instantly see a rough pre-vis, or software that handles the tedious parts of color grading so the colorist can focus on creative choices. On the more contentious end: AI-assisted script analysis, automated VFX generation, or synthetic voice work. The Writers Guild and SAG-AFTRA spent much of 2023 and 2024 negotiating AI protections into their contracts. Those agreements set guardrails, but they did not ban AI from production workflows. The door is open. The question is who walks through it first, and with what intentions.

Some filmmakers are already experimenting at the edges. Recent tools from Runway and Adobe have made AI-assisted rotoscoping, background replacement, and shot matching dramatically faster than manual methods. DeepMind’s own Veo video generation model has been demoed generating short clips from text prompts, though it has not been released for commercial use. The step from “generate a 10-second clip of a sunset” to “generate a rough assembly of scene 47 so the director can see if the blocking works” is not very large. The step from that to “generate the final shot” is larger — and more contested. A24’s role in this partnership, at least on paper, is to make sure the tools stay on the first side of that line unless the filmmaker explicitly chooses otherwise.

A24’s involvement might actually constrain the worst impulses. A studio that built its reputation on authorial voice — on Ari Aster’s singular vision, on the Daniels’ anarchic creativity — has every incentive to ensure the tools amplify individual directors rather than replace them. If A24’s filmmakers come back and say “this feature is useless” or “this tool produced something generic,” DeepMind has to listen. That is the premise of the partnership.

The broader race is already underway

The A24 deal did not happen in isolation. Netflix’s acquisition of InterPositive — Ben Affleck’s company, which had been developing AI tools specifically for filmmakers — closed at $587 million. That is more than seven times what Google is putting into A24, and it gave Netflix ownership of the technology rather than a partnership stake. Amazon MGM’s AI unit has been operational since early 2025, building internal tools for television and film production. Apple and Disney have both made AI hires in their production divisions, though neither has announced a formal studio partnership yet.

What is emerging is a two-tier structure. On one tier, vertically integrated streamers like Netflix and Amazon are buying or building AI capabilities outright, keeping the technology in-house. On the other tier, Google is taking a partnership approach — investing in a studio rather than acquiring one, building tools collaboratively rather than imposing them. Neither approach is obviously better or worse for filmmakers. But they lead to different power dynamics.

In the Netflix model, the tools belong to the platform. Directors who make films for Netflix use Netflix’s tools under Netflix’s terms. In the Google-A24 model, the tools could theoretically be made available more broadly — Google’s core business is selling cloud services and AI infrastructure, not producing movies. DeepMind’s incentive is to prove its AI works in a creative context. A24’s incentive is to give its filmmakers an edge. If the tools work, other studios will want them too, and Google benefits.

What independent filmmakers should watch for

The most important unknown is whether any of this changes things for filmmakers outside the A24 orbit. If DeepMind develops a pre-visualization tool that a director can use on a laptop, does that tool become available through Google Cloud with a per-seat license? Or does it remain exclusive to A24 productions as a competitive advantage? Google has not said, and the ambiguity is probably intentional — keeping options open while the tools are still being defined.

The stakes are higher than they might seem. Independent film financing has been getting harder, not easier. Mid-budget films — the $5 million to $30 million range where most of A24’s catalog lives — have been squeezed by the streaming era. If AI tools can meaningfully reduce production costs without compromising what makes those films distinctive, the partnership could produce tools that genuinely help the filmmakers who need them most. But that only works if the tools escape the walled garden.

For independent filmmakers — the ones shooting on borrowed locations with crews of eight — the difference between “A24 gets a cool new tool” and “anyone with a Google account can access a cool new tool” is the difference between widening the gap and closing it. The history of technology in filmmaking leans toward democratization. Digital cameras put film-quality images within reach of filmmakers who could not afford 35mm stock. Editing software turned bedroom laptops into cutting rooms. If AI tools follow that pattern, the A24-DeepMind partnership could eventually produce tools that help a first-time director in Cleveland pre-visualize a scene as effectively as a studio production.

But AI also has a history of concentrating power. The most advanced models require computing resources that individuals do not have. If the best AI filmmaking tools remain tethered to Google’s cloud infrastructure, the gap between well-funded productions and everyone else could widen even as the tools themselves become more capable.

The real test is the first film

For now, the entire conversation is speculative. The partnership has been announced. The money has changed hands. The tools do not exist yet — at least not in any form the public has seen.

The real test will be the first film that credits DeepMind tools in its production notes. If that film looks like something only A24 would make — strange, personal, unwilling to sand down its edges — then the partnership will have proven its thesis. If it looks like something an algorithm optimized for engagement metrics, the industry will have its answer about which partner set the terms.

A24 has spent a decade building a reputation as the studio that protects directors from the machinery that makes most movies feel the same. Google just handed them $75 million and asked to be let inside the tent. The next year or two will tell us whether that was an invitation to collaborate or a very expensive Trojan horse.

Share This Article

Comments

Join the conversation

0 entries

Loading comments...

Related Articles