Two Hollywood stories broke in the same few days, and read together they say more than either does alone. A24, the studio whose whole brand is being the tasteful alternative to algorithmic content, announced a research partnership with Google DeepMind to build production tools. Meanwhile, in a Los Angeles courtroom, the major studios suing the AI image generator Midjourney were fighting to keep their own use of AI out of evidence. One industry, two postures: enthusiastic adopter and outraged victim, depending on which way the camera is pointing.
The A24 deal, reported first by The Wall Street Journal and covered in detail by CBR, is being sold as artists getting a seat at the table. DeepMind's product chief framed it as anchoring the technology in the creative process, and A24 told Wired it would rather help shape the tools than have them handed down. The stated guardrails are real enough: the tools will not be trained on A24's film library, and the studio says it does not love current AI output on screen. Fans were unconvinced, brigading the announcement and cancelling memberships, and the sceptics have a point about how quickly a workflow experiment can become a house style.
What CBR draws out is that A24's ties run deeper than one research agreement. The studio's 2024 funding came from Thrive Capital, an early OpenAI backer whose founder now sits on A24's board, and which also holds a stake in the model-evaluation firm Scale AI. None of that means A24 is quietly building a database of actors' likenesses. It does mean the wall between prestige filmmaking and the generative-AI industry is thinner than the studio's image suggests, and that the normalising of these tools behind the scenes is the actual story, more than any single clip they might produce.
The Midjourney case, reported by TechCrunch, is where the double standard becomes explicit. Disney, Universal, and Warner Bros. have sued the startup for generating images of characters they own, from Bart Simpson to Darth Vader. Midjourney's countermove is pointed: it is trying to compel the studios to disclose how they use AI internally. If the studios are training their own image models on unlicensed material for storyboarding or ideation, Midjourney argues, that would show the very practice they are suing over is industry custom, even among the plaintiffs.
A judge has so far limited the studios' disclosure to AI use that reached consumer-facing work, and Midjourney wants that limit lifted, along with every prompt and output the studios have run through its own service. The studios call it a fishing expedition and insist they only want their characters left alone. Both things can be true. The request can be a litigation tactic and still point at something uncomfortable: the companies loudest about copyright may be relying on the same shortcuts when no one is watching.
The throughline is not hypocrisy so much as leverage. Studios want AI on their terms, as a private efficiency they control, not as a public product that lets anyone conjure their characters for free. A24's deal is an attempt to own the tools before the tools own the workflow. The lawsuits are an attempt to police the boundary of who gets to copy whom. What this week clarified is that Hollywood is not against generative AI. It is against everyone else having it.