Midjourney Appeals Order Limiting AI Use Disclosure
Midjourney is asking a federal judge in the Central District of California to overturn a magistrate judge's June 16, 2026 order that limited its discovery into how Disney, Universal, and Warner Bros. use AI internally, in the consolidated copyright lawsuits the studios brought over AI-generated Batman, Superman, and other characters. Midjourney argues the studios' own AI-training practices are directly relevant to its fair-use and unclean-hands defenses, and that the magistrate judge wrongly drew a line between the studios' public-facing AI tools and their internal AI use. The studios had resisted broader discovery; the magistrate allowed only some requested data through while shielding the rest as irrelevant or protected work product. The outcome could set precedent for how much visibility AI companies get into plaintiffs' own AI practices during copyright discovery.
For teams tracking AI copyright litigation, this discovery fight previews a defense strategy spreading across AI lawsuits: model makers accused of training on copyrighted works are pushing courts to let them examine whether the plaintiffs suing them train their own AI systems the same way, turning a copyright holder's AI practices into evidence for the defendant.
What happened
Midjourney has asked the district judge overseeing its case to review and reverse a June 16, 2026 order from the magistrate judge (Richlin) that limited Midjourney's discovery requests into Disney's, Universal's, and Warner Bros.' internal use of artificial intelligence. The underlying lawsuits, filed in the U.S. District Court for the Central District of California (including Disney Enterprises v. Midjourney, case 2:25-cv-05275, and Warner Bros. Entertainment v. Midjourney, case 2:25-cv-08376), accuse Midjourney of enabling users to mass-generate copyrighted characters such as Batman and Superman. The magistrate judge's order let the studios withhold broader data on their AI use, finding some of Midjourney's requests irrelevant or shielded by work-product privilege, while requiring production of some narrower categories.
Background
According to Mealey's, which reviewed Midjourney's filing, Midjourney's motion argues the evidence goes to its fair-use and unclean-hands defenses, and that there is no principled basis for distinguishing between the studios' public-facing AI products and their internal AI training and tooling. If Midjourney can show the studios train their own models on copyrighted works in ways similar to what Midjourney is accused of, it could weaken the studios' claim of harm or support an unclean-hands defense that limits Midjourney's liability. Law360's case records list U.S. District Judge John A. Kronstadt as presiding over the matter.
For practitioners
This dispute sits inside a discovery pattern showing up across AI copyright litigation: defendant AI companies are seeking plaintiffs' own AI training records as leverage, testing whether an everyone-trains-on-copyrighted-data argument can survive discovery scrutiny rather than just serve as rhetoric. Legal and policy teams tracking AI training-data exposure should watch how courts balance relevance against burden and privilege claims here, since the ruling could shape discovery scope in the many other AI copyright suits working through federal courts.
What to watch
The district judge's ruling on Midjourney's motion for review will determine whether the studios must produce broader internal AI-use data. A win for Midjourney would set a notable precedent expanding discovery into plaintiffs' own AI practices in copyright suits against AI companies; a loss would reinforce current limits on this kind of discovery strategy.
Key Points
- 1Midjourney asked a federal judge to reverse Magistrate Judge Richlin's June 16, 2026 order limiting discovery into Disney's, Universal's, and Warner Bros.' internal AI use.
- 2Midjourney argues the studios' own AI-training practices bear directly on its fair-use and unclean-hands defenses in the underlying copyright lawsuits.
- 3The ruling on review could set precedent for how much AI-training discovery courts allow AI-company defendants to obtain from copyright-holder plaintiffs.
Scoring Rationale
Notable procedural development in the marquee studio-vs-Midjourney copyright litigation: the discovery fight over the studios' own AI training practices could shape fair-use and unclean-hands defenses and set a template for discovery scope across dozens of parallel AI copyright suits. Score reflects real precedential stakes, tempered by the fact this is an interim discovery ruling rather than a decision on the merits.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
