The Midjourney Hollywood studios discovery dispute has taken a sharper turn, with the AI image-generation startup filing a motion to strip away limits on what Disney, Universal, and Warner Bros. must disclose about their own artificial intelligence activity.
The three studios collectively sued Midjourney for alleged copyright infringement, with Georgetown Law’s Tech Institute noting that Disney, Universal, and DreamWorks Animation filed a 110-page complaint on 11 June 2025 in the U.S. District Court for the Central District of California. Warner Bros. followed with its own action, and on 4 November 2025 the two cases were consolidated by the court. The studios’ complaint, which characterised Midjourney as a ‘bottomless pit of plagiarism’ and a ‘virtual vending machine for unauthorised copies,’ centres on claims that the startup’s models can generate recognisable likenesses of characters including Bart Simpson and Darth Vader.
Midjourney argues its training practices fall under fair use. Now it wants the studios to prove they are not doing the same thing behind closed doors.
Midjourney Hollywood Studios Discovery Fight Turns on the Studios’ Own AI Conduct
Magistrate Judge Joel Richlin had previously ruled that the studios needed to produce documents related to their generative AI usage, but only where that usage led to consumer-facing videos and images. According to VP-Land’s reporting on the motion, Midjourney’s latest filing seeks to overturn that limitation, arguing that the judge’s ruling permits the studios to ‘cherry-pick only those documents they believe support their market harm claims while depriving Midjourney of documents that would support its defences.’
The startup is asking for a broad range of internal materials: AI business plans, research reports, training data, model weights, and board presentations on AI strategy. It also wants every prompt the studios have entered into Midjourney, together with the resulting outputs, not just those prompts that allegedly produced infringing images.
The logic is direct. If the studios are developing image-generating models for internal use in storyboarding or developing content for film and television, Midjourney contends that would ‘demonstrate that it is an industry custom, even among the studios themselves, to download and train AI on unlicensed copyrighted content.’
Midjourney’s attorneys went further in the filing, arguing that if the studios are ‘doing the very thing they seek to punish, that evidence goes to the heart of Midjourney’s fair use and unclean hands defences,’ according to a quote attributed to attorney Ghajar in the filing.
The Studios’ Own Words Create Complications
The studios are not uniformly hostile to AI, which complicates their posture. In a 43-page response filed in the same court, Midjourney’s attorneys quoted Disney chief executive Bob Iger as stating that ‘technology is an invaluable tool for artists,’ adding that generative AI is ‘no different,’ according to Artnet News. Disney’s senior executive vice president and chief legal and compliance officer, Horacio Gutierrez, said at the time of the lawsuit filing that the company is ‘bullish on the promise of A.I. technology and optimistic about how it can be used responsibly as a tool to further human creativity.’
The studios’ lead attorney, David Singer, has pushed back on the scope of Midjourney’s requests, describing them as a ‘fishing expedition.’ He has also stated that the studios ‘do not seek to stop AI technology or even shut down Midjourney’s business,’ 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.’
Midjourney’s own record on this question is not clean. In a 2022 interview, founder David Holz acknowledged that the company pulls text and images from the internet to collect training data, admitted it did not seek consent from copyright holders, and said it was open to an opt-out model for rights holders, though no such process was in place, according to Stone Law’s analysis of the complaint. The studios’ 110-page complaint also alleged that Midjourney failed to respond to pre-suit letters notifying it of the infringing activity.
This litigation carries broader stakes. As Simon Willison’s review of the case points out, it marks the first time major Hollywood studios have brought a copyright action against an AI image generator; earlier AI copyright litigation had included a class action brought by visual artists against Midjourney in 2023.
The discovery battle now hinges on whether Judge Richlin’s consumer-facing limitation survives Midjourney’s challenge. If the startup succeeds, the studios could be forced to open their internal AI programmes to scrutiny precisely as their market-harm argument is being constructed. That ruling, whenever it comes, will set the evidentiary terms for every fair-use argument in the case.
