Midjourney is trying to turn a copyright lawsuit into a broader examination of how major Hollywood studios use generative AI. In a new filing, the AI startup argues that Disney, Universal and Warner Bros. should be required to disclose more about their own AI practices as the case moves through discovery.
What the studios allege
Disney and Universal sued Midjourney for alleged copyright infringement last year. Their complaint focused on the startup's image-generation models and their ability to produce images of characters owned by the studios, including Bart Simpson and Darth Vader.
A few months later, Warner Bros. sued Midjourney as well. Together, the cases place the startup's AI training practices and outputs at the center of a dispute over copyrighted entertainment properties.
Midjourney's position is that training AI models on images of copyrighted characters is allowed under fair use. That argument is now shaping what the company wants to obtain from the studios during discovery.
The discovery fight
The immediate dispute is not over the final copyright question. It is about what documents the studios must turn over.
A judge previously ruled that the studios did have to provide information about their generative AI usage. But that order was limited to cases where the AI use resulted in “consumer-facing” videos and images.
Midjourney now wants that limitation removed. In its latest filing, the company says the current boundary “unfairly” allows the studios “to cherry-pick only those documents they believe support their market harm claims while depriving Midjourney of documents that would support its defenses.”
That is the core of Midjourney's argument: if the studios are allowed to define the relevant AI records too narrowly, the startup says it may be blocked from evidence that could help its case.
Why internal AI use matters to Midjourney
Midjourney is especially interested in AI activity that may never have reached the public. The company argues that internal experimentation, development and creative work could still be relevant to whether its own practices fit within accepted industry behavior.
In the filing, Midjourney claims 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 gives one example: studio development of image-generating AI models “for internal use in storyboarding or ideating content for film or TV.” Midjourney says evidence of that kind would support an argument that it is an industry custom, including among the studios themselves, “to download and train AI on unlicensed copyrighted content.”
That argument does not resolve the copyright dispute by itself. But it explains why Midjourney wants records that go beyond finished public videos and images. The company is trying to connect the studios' private AI use to its defense against claims that its own model training and image generation caused legal harm.
Prompts are also in dispute
Midjourney is also asking for a broader set of prompt records. According to the filing, the startup wants the studios to reveal all the prompts they used in Midjourney, along with the outputs those prompts generated.
The studios have focused on prompts that produced the allegedly infringing images. Midjourney says that is too narrow. It wants the full prompt-and-output picture, not only examples selected because they support the studios' infringement claims.
That distinction matters because prompts can show how a user steered the system. Outputs can show what resulted from those instructions. Together, they may help define whether the contested images reflect the model's behavior, the user's choices, or both.
The studios push back
The studios' lead attorney David Singer previously described Midjourney's request for this documentation as a “fishing expedition.”
He also said the studios “do not seek to stop AI technology or even shut down Midjourney’s business.” Instead, he said they “simply want Midjourney to stop copying their movies and TV shows and to stop distributing, publicly displaying, publicly performing, and creating derivative works that include copies of [their] famous characters without authorization.”
That framing keeps the studios' argument focused on their characters, movies and TV shows, rather than on generative AI as a technology category. Midjourney's filing pushes in the opposite direction by making the studios' own generative AI use part of the factual record it wants the court to consider.
The result is a discovery fight with larger implications for AI and entertainment. The court is being asked to decide how much of the studios' AI activity Midjourney can inspect while defending itself against claims involving copyrighted characters and alleged infringement.