Midjourney asks court to force Hollywood studios to reveal their own AI training practices

In a significant escalation of the ongoing legal battle between AI startup Midjourney and three major Hollywood studios, the company has filed a motion seeking to compel Disney, Universal, and Warner Bros. to disclose detailed information about their own internal use of generative artificial intelligence. The filing argues that the studios should not be allowed to sue Midjourney for copyright infringement while potentially engaging in similar practices themselves.

Background of the dispute

Disney and Universal filed a lawsuit against Midjourney in 2024, alleging that the startup’s image-generation models could produce unauthorized copies of iconic copyrighted characters such as Bart Simpson and Darth Vader. Warner Bros. later joined the suit. Midjourney has maintained that training its AI models on publicly available images, including those featuring copyrighted characters, falls under the fair use doctrine — a legal principle that permits limited use of copyrighted material without permission under certain circumstances.

Discovery dispute over AI documentation

The current conflict centers on the scope of documentation the studios must produce during the discovery phase of the litigation. A judge previously ruled that the studios must provide information about their generative AI usage, but only when that usage resulted in “consumer-facing” videos and images. Midjourney is now seeking to overturn that limitation, arguing that it unfairly allows the studios to selectively produce documents that support their claims of market harm while withholding evidence that could support Midjourney’s defense.

Midjourney’s central argument

In its latest filing, Midjourney contends that the documents the studios are withholding “are precisely those that would reveal whether, behind closed doors, they are doing exactly what they are suing Midjourney for doing.” The startup argues that if the studios are developing image-generating AI models for internal purposes such as storyboarding or ideating content for film or television, that evidence would demonstrate that downloading and training AI on unlicensed copyrighted content is an industry custom — even among the studios themselves.

Scope of requested information

Beyond internal AI training practices, Midjourney is also seeking disclosure of all prompts the studios used in its platform, along with the resulting outputs — not just the prompts that allegedly produced infringing images. The startup argues that a complete record is necessary to assess the full context of the studios’ claims and to evaluate any potential hypocrisy in their litigation strategy.

Studios’ response and characterization

David Singer, lead attorney for the studios, has described Midjourney’s request as a “fishing expedition.” Singer has emphasized that the studios “do not seek to stop AI technology or even shut down Midjourney’s business,” but rather want the startup to stop copying their movies and TV shows and to cease distributing derivative works that include unauthorized copies of their famous characters.

Why this matters

This case has broader implications for the generative AI industry and the entertainment sector. If Midjourney succeeds in compelling the studios to disclose their internal AI practices, it could set a precedent for transparency in AI training data and usage across the industry. Conversely, if the court upholds the limitation on discovery, it may limit the ability of AI companies to challenge copyright claims by pointing to similar conduct by the plaintiffs. The outcome could influence how both technology companies and content creators approach the legal boundaries of AI training and fair use.

Conclusion

The dispute between Midjourney and the Hollywood studios represents a pivotal moment in the intersection of copyright law and generative AI. As the court considers whether to expand the scope of discovery, the decision will likely shape the legal landscape for AI training practices and the responsibilities of companies that both create and use copyrighted content. The case continues to unfold, with both sides preparing for what could be a lengthy and consequential legal battle.

FAQs

Q1: What is Midjourney asking the court to do?
Midjourney is asking the court to compel Disney, Universal, and Warner Bros. to disclose detailed information about their own internal use of generative AI, including training data and prompts used on Midjourney’s platform, not just consumer-facing outputs.

Q2: Why does Midjourney argue this information is relevant?
Midjourney argues that if the studios are also training AI on unlicensed copyrighted content for internal purposes, it would demonstrate that such practices are an industry custom and could support Midjourney’s fair use defense.

Q3: What is the current status of the discovery dispute?
A judge previously ruled that the studios must provide information about AI usage leading to consumer-facing content. Midjourney is now seeking to overturn that limitation and require broader disclosure.