A24 has built its reputation around distinctive filmmakers and unconventional creative choices. That identity is precisely what makes its new partnership with Google DeepMind both consequential and controversial.<\/p>\n
Announced on June 22, the agreement brings one of the film industry\u2019s most influential independent studios into a long-term research collaboration with Google\u2019s artificial intelligence laboratory. The stated aim is to place emerging technology directly within the creative process, allowing filmmakers to influence the development of new tools and production methods.<\/p>\n
Yet the announcement arrives at a moment when artificial intelligence remains one of the industry\u2019s most contested subjects. Questions surrounding copyright, consent, employment and creative control have moved from speculative debate into union agreements, courtrooms and active production policy.<\/p>\n
Inside the A24\u2013DeepMind Agreement<\/p>\n
According to the official announcement<\/a>, the partnership will span multiple projects, with DeepMind researchers working alongside A24 and its filmmakers to test and develop new workflows.<\/p>\n Google has also made an investment in the studio. The Wall Street Journal reported the amount at approximately $75 million<\/a>, describing it as Google\u2019s first equity investment in a film company.<\/p>\n The known limits of the deal are equally important:<\/p>\n Non-Exclusive R&D: It is a multiyear research partnership, meaning A24 remains free to collaborate with other technology entities.<\/p>\n Library Protections: Google will not receive access to the studio\u2019s film and television library or underlying catalog data.<\/p>\n No Active Model Training: There is no indication that A24 intends to train a proprietary model on its existing catalog or deploy unvetted tools into active productions immediately.<\/p>\n For now, the arrangement is strictly R&D: A24 provides direct access to real-world creative workflows and filmmaker feedback, while DeepMind supplies research expertise, compute infrastructure, and technical development.<\/p>\n Industry Pushback & Creative Control<\/p>\n The distinction between research and production did not prevent a backlash. A24\u2019s audience has long associated the company with filmmaker-led work and a resistance to the safest conventions of mainstream studio filmmaking. For some supporters, partnering with one of the world\u2019s largest technology companies appeared fundamentally at odds with that image.<\/p>\n A24 responded by arguing that it would rather have a \u201cseat at the table\u201d than allow creative technology to be designed without filmmakers. Scott Belsky, who oversees the studio\u2019s technology initiative, has said the planned tools should not resemble the prompt-based content generation that has attracted much of the criticism surrounding generative AI.<\/p>\n The studio\u2019s position is that working directly with DeepMind could help preserve creative control. Critics, however, remain concerned that tools initially presented as creative assistance can later become instruments for reducing costs, compressing schedules or replacing parts of established production work.<\/p>\n That concern is not theoretical. SAG-AFTRA\u2019s newly ratified 2026 TV\/Theatrical Agreement expanded protections around AI, synthetic performers and digital replicas, reflecting how central consent and human performance have become to current labor negotiations. The agreement was approved in June<\/a>, with 91.42% of participating members voting in favor.<\/p>\n At the same time, major studios are pursuing copyright cases against AI companies accused of using protected material without permission. A24\u2019s decision to keep its own library outside the DeepMind agreement therefore represents a meaningful boundary\u2014but not a complete answer to how future tools will handle training data, ownership, attribution or consent.<\/p>\n Where AI could enter the production process<\/p>\n Artificial intelligence does not have to generate actors or finished scenes to influence how a film is made. Its more immediate applications could emerge around the image rather than inside it.<\/p>\n Production planning, script breakdowns, scheduling, previsualization, asset searches, temporary visual effects, localization and post-production management are all areas in which machine-assisted systems could alter existing workflows. None of these applications has been confirmed as part of the A24 programme, but they illustrate the range of decisions covered by the broad language of \u201cnew workflows and techniques.\u201d<\/p>\n The practical questions will therefore concern more than visual quality. Productions will need to establish who approved the use of a tool, what material it processed, where that data was stored and whether its output influenced work that would otherwise receive a human credit.<\/p>\n The most significant effect may initially be procedural. New systems could change how departments exchange information, how early creative decisions are visualized and how production data travels between studios, vendors and post-production partners.<\/p>\n Why Budapest will be watching<\/p>\n A24\u2019s relationship with Hungary gives these questions a tangible production context. The studio released Death of a Unicorn, starring Paul Rudd and Jenna Ortega, which was shot entirely in Hungary<\/a>.<\/p>\n