Midjourney Demands Disclosure of Hollywood Studios' Own AI Use

Midjourney, the popular AI image generator, has filed a motion in its ongoing copyright infringement case demanding that Disney, Universal and Warner Bros. turn over detailed documentation of their own internal use of artificial intelligence. It's the latest twist in a legal battle that has been running since 2025.

The studios argue that Midjourney's models generate unauthorized depictions of their copyrighted characters, infringing their rights. Midjourney counters that training its models on copyrighted images falls within fair use, and accuses the studios of presenting evidence selectively.

What Midjourney is demanding

Midjourney's motion seeks complete records of the studios' internal development of AI tools, including storyboarding and concept-creation applications, every prompt entered into Midjourney itself (not just those that produced the disputed images), and the resulting output. The company wants to show how the studios themselves use generative AI behind closed doors.

Midjourney's argument rests on the claim that the studios are applying a double standard: suing the company for training its models on copyrighted material while using similar AI tools for their own production purposes without disclosing the practice publicly.

The studios' position

David Singer, the attorney representing the studios, framed the purpose of the lawsuit in simpler terms, rejecting accusations of inconsistency.

We just want them to stop copying our movies and TV shows - David Singer, attorney representing the studios

For the studios, what matters is that Midjourney's models can generate recognizable images of characters like Bart Simpson or Darth Vader without any license, which in their view constitutes direct copyright infringement regardless of how they themselves use similar technology internally.

Industry implications

The case touches on one of the most contested questions in copyright law in recent years: whether training generative models on copyrighted material falls within fair use. The outcome of this particular clash could affect other pending lawsuits against AI companies, including disputes over AI-generated music and text.

The court must now decide whether to lift the current restriction that has limited disclosure to documents related to the studios' consumer-facing content. Expanding the scope of disclosure to internal production tools would set a precedent that could force other major media companies into greater transparency in disputes with AI vendors.

For Polish creators and companies using generative AI in content production, the case shows that the line between fair use and copyright infringement remains far from settled even in the United States, where most of these precedent-setting lawsuits are playing out.

Sources: TechCrunch (techcrunch.com)