Midjourney to Hollywood: 'Show Us Your AI Usage, Too'

Midjourney is flipping the script on the studios that sued it. In a legal maneuver filed last week, the AI image generation company is asking a court to compel Disney, Universal, and Warner Bros. to disclose their own internal AI usage — turning what started as a straightforward copyright infringement case into a transparency standoff the studios probably didn’t see coming (TechCrunch, July 4, 2026).

The Case So Far

Disney and Universal sued Midjourney last year, arguing that the startup’s image-generation models could produce images of characters like Bart Simpson and Darth Vader — intellectual property the studios own. Warner Bros. filed a similar suit a few months later. The core argument is familiar: AI models trained on copyrighted material can reproduce it, and the companies behind those models should be liable.

Midjourney’s response takes the case in an unexpected direction. Rather than solely defending its training practices, the company is demanding discovery into how the studios themselves use generative AI. The implication is hard to miss: if Hollywood is deploying AI tools across production, visual effects, marketing, and concept art — which it almost certainly is — then the studios are simultaneously suing over AI-generated images while using AI to generate images.

Why This Matters for the AI Art Industry

The Midjourney strategy isn’t just legal theater. It exposes a tension that’s been building since the first AI copyright lawsuits were filed: the companies most aggressive about protecting their IP from AI are often the same companies quietly integrating AI into their pipelines.

Several major studios have publicly acknowledged experimenting with generative AI for concept art, storyboarding, and post-production. Disney’s own research division has published papers on AI-assisted animation. None of that is illegal, but it complicates the narrative that AI image generation is inherently infringing when the plaintiffs are using similar technology.

If Midjourney’s discovery request succeeds, the internal documents that surface could reshape the broader AI copyright debate. Not because they’ll prove or disprove infringement, but because they’ll show that the line between “AI artist” and “AI defendant” is blurrier than either side has been willing to admit.

This isn’t Midjourney’s only legal fight. It is also a defendant in Andersen v. Stability AI, a proposed class action brought by artists. But the Hollywood case is different because the plaintiffs aren’t just rights holders — they’re active technology adopters.

Case timeline (from court records and reporting)

• June 11, 2025: Disney and Universal sued Midjourney for copyright infringement in the U.S. District Court for the Central District of California (Reuters). The case is Disney Enterprises, Inc. v. Midjourney, Inc., No. 2:25-cv-05275 (CourtListener docket).
• September 4, 2025: Warner Bros. Discovery filed a similar complaint. The parties later stipulated to consolidate the two actions before Judge John A. Kronstadt.
• June 2026: Magistrate Judge Joel Richlin limited Midjourney’s discovery into the studios’ AI use, ordering production only about consumer-facing AI applications and finding broader requests irrelevant or protected (Variety; Law360).
• Following that ruling: Midjourney asked Judge Kronstadt to overturn the limit, arguing that the studios’ internal AI use is relevant to its fair-use and “unclean hands” defenses.

Why discovery matters here

In U.S. federal litigation, discovery lets each side obtain evidence relevant to claims and defenses, subject to proportionality and privilege limits. Midjourney’s argument is that if studios themselves train or use image models on unlicensed material, that bears on industry custom and fair use. The studios’ position, accepted in part by the magistrate judge, is that their own practices do not decide whether Midjourney infringed. How the district judge resolves this will signal how far AI defendants can probe plaintiffs’ AI use in similar cases.

Related cases to watch

Midjourney is not the only image-model company in court. Andersen v. Stability AI, brought by a group of artists, names Stability AI, Midjourney, DeviantArt, and Runway, and is pending in the Northern District of California. Separately, Getty Images has sued Stability AI (not Midjourney) in the U.S. and the U.K. Outcomes in these cases will shape licensing norms for training data.

What Comes Next

In June 2026, a magistrate judge limited the discovery, ordering the studios to produce information only about consumer-facing AI applications; Midjourney has asked the district judge to overturn that limit. If the broader request is granted, the studios would have to produce records of internal AI tool usage, training data practices, and policies. If it’s denied, the case proceeds largely on the original copyright claims.

For independent AI artists and smaller studios watching the case, the outcome matters in practical terms. A ruling that favors the studios on copyright grounds but allows them to keep their own AI use opaque creates a two-tier system: big companies get to use AI however they want, while smaller players face legal exposure for the same tools. That’s not a hypothetical — it’s already the dynamic in music, publishing, and visual effects. Midjourney’s discovery motion is an attempt to make that double standard visible in court records, not just industry gossip.