Google DeepMind Invests in A24 for AI Research

Google DeepMind and independent film studio A24 announced a multiyear research collaboration, according to Google DeepMind's blog post and reporting by Reuters. The Wall Street Journal and trade publications report that Alphabet's Google is investing approximately $75 million in A24 as part of the initiative. Google DeepMind describes the collaboration as research-focused and aimed at helping artists develop new production workflows and techniques.

Variety quotes A24 partner Scott Belsky, who oversees the studio's technology and innovation work, saying, "We think there are better uses [of AI] that preserve creative control and support risk-taking," and quotes DeepMind VP Eli Collins as saying, "We believe breakthroughs happen when you get technology into the hands of the best minds in the field" (Variety). Reuters reports a person familiar with the matter told the outlet that the arrangement is not an intellectual-property or data-training deal and that filmmakers would retain full creative control.

Google DeepMind's public announcement does not list specific technical outputs, models, or datasets. Multiple outlets, including The Verge and Reuters, describe the collaboration as spanning "multiple projects over time" and as non-exclusive. Reuters reports the partnership will provide A24 with access to DeepMind research and infrastructure, per a source familiar with the matter. A24 Labs, Belsky's 20-person tech team, has early projects underway including using AI to generate storyboards, per TechCrunch.

Industry observers note that formal research partnerships between large AI labs and creative studios typically focus on prototyping tools for asset management, content-aware editing, simulation, and workflow automation, while leaving final creative decisions to human authors. Such collaborations can produce technical work in areas like conditional generation, style transfer, and multimodal orchestration; practitioners designing film-focused systems often face elevated requirements for provenance, versioning, and fine-grained access control. Reuters and The Verge both emphasize the non-exclusive, non-data-training nature of the arrangement, which is unusual in a landscape where studios have resisted broad AI training grants.

Key indicators include whether specific tools or prototypes emerge from the collaboration, how A24 discloses the partnership on credits or press materials, and whether similar arrangements proliferate among other independent studios. Practitioners in creative AI should monitor whether the research outputs are published or remain proprietary to the studios involved.