AI image generator Midjourney is trying to turn the tables on the Hollywood studios suing it for copyright infringement. In a new court filing, the startup is asking a judge to force Disney, Universal, and Warner Bros. to disclose how they themselves use generative AI, arguing the studios may be doing the very thing they are accusing Midjourney of doing.
The dispute has become a revealing early test of how the US legal system will weigh corporate AI practices against intellectual property claims. For ongoing breaking AI news on the litigation reshaping generative AI, the discovery phase of this case could set important precedents for what companies must reveal about their internal AI tools.
The Lawsuits Behind the Fight
Disney and Universal sued Midjourney in June 2025, alleging that the startup's image-generation models could produce unauthorized likenesses of copyrighted characters such as Bart Simpson and Darth Vader. A few months later, in September 2025, Warner Bros. filed its own lawsuit over images of characters including Superman and Batman.
At the core of each complaint is the same accusation: that Midjourney trained its models on copyrighted material without permission and then allowed users to generate near-identical copies of famous characters. The studios contend this amounts to wholesale infringement of their most valuable intellectual property.
The current skirmish, however, is not about whether infringement occurred. It is about the discovery process, the pretrial exchange of evidence, and specifically what the studios must hand over about their own use of generative AI.
The Discovery Dispute
A judge had previously ruled that the studios would need to provide information about their generative AI usage, but only when it led to "consumer-facing" videos and images. That limitation meant internal or behind-the-scenes AI work would stay hidden.
Midjourney wants that restriction overturned. In its latest filing, the startup argues the consumer-facing carve-out "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."
Midjourney goes further, suggesting the studios may be quietly relying on the same practices they condemn. The documents being withheld, the filing states, "are precisely those that would reveal whether, behind closed doors, they are doing exactly what they are suing Midjourney for doing."
An Industry Custom Argument
The filing lays out a pointed hypothetical. If the studios are developing image-generating AI models for internal use, such as storyboarding or ideating content for film or television, Midjourney argues that evidence would show "it is an industry custom, even among the studios themselves, to download and train AI on unlicensed copyrighted content."
That framing is strategically significant. Copyright law looks at factors including whether a use is widespread and whether it harms the market for the original work. If Midjourney can show that the studios themselves routinely train AI on copyrighted material, it weakens the argument that Midjourney's conduct is exceptional or that the studios are being uniquely harmed.
Midjourney is also seeking a broader set of records than the studios want to produce. The startup argues the studios should reveal all of the prompts they entered into Midjourney and the resulting outputs, not merely the prompts that generated the allegedly infringing images. Midjourney's position is that the full record is needed to understand how the disputed images were actually created and used.
The Studios Push Back
The studios have dismissed Midjourney's demands as overreach. Their lead attorney, David Singer, previously characterized the request for documentation as a "fishing expedition."
Singer has also sought to frame the studios' position as measured rather than anti-technology. The studios "do not seek to stop AI technology or even shut down Midjourney's business," he said, but rather "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 distinction matters as the case moves forward. The studios are drawing a line between using AI as a tool and using it to reproduce protected characters, while Midjourney is pressing to expose any inconsistency between what Hollywood preaches about AI and what it practices internally.
Why the Outcome Matters
The Midjourney case sits at the center of a broader wave of copyright litigation against generative AI companies, and the discovery rules that emerge here could ripple across the industry. If courts require plaintiffs who sue AI companies to disclose their own AI usage, it could complicate future claims from rights holders who themselves depend on machine-learning tools in their production pipelines.
The fight also highlights a tension at the heart of Hollywood's relationship with AI. Studios have publicly criticized unauthorized training while simultaneously investing in and experimenting with generative tools for their own creative work. How much of that internal experimentation must be revealed is now a question for a federal judge.
As discovery continues, both sides are effectively asking the court to define the boundaries of acceptable AI use in a creative industry that is simultaneously building and battling the technology.
Stay Ahead of AI
Copyright disputes are quietly defining the rules for the entire generative AI industry. Keep up with the AI industry coverage that tracks these cases as they unfold.
Read more AI news →
