One of the biggest legal battles in the AI industry has taken an unexpected turn. Instead of simply defending itself, Midjourney is now asking the court to require Disney, Warner Bros. Discovery, and Universal Pictures to disclose how they have been using artificial intelligence behind the scenes.
The company is seeking access to AI prompts entered by the studios’ employees, the images generated through Midjourney, and even internal AI-generated materials that were never intended for public release.
How did it all start?
The dispute began in 2025, when Disney and Universal sued Midjourney, arguing that its image generator could create highly recognizable versions of iconic characters from simple text prompts. A few months later, Warner Bros. joined the lawsuit.
The studios claim Midjourney is unlawfully reproducing copyrighted characters without permission. Under U.S. copyright law, they are seeking up to $150,000 in damages for each infringed work, potentially exposing Midjourney to massive financial liability. Midjourney, however, maintains that training AI models on publicly available images falls under the legal principle of fair use.
Earlier in the case, the court ordered the studios to provide limited information about their use of generative AI—but only for AI-generated content that was ultimately released to audiences. Midjourney argues that this limitation should be removed.
The company wants access to internal AI-related materials as well, including storyboards, concept art, creative drafts, and other content generated during production, even if it never reached the public.
According to the company, the documents could reveal how Hollywood studios themselves use AI. If Disney, Warner Bros., or Universal have trained or developed AI systems using copyrighted material for internal creative work, Midjourney believes it would demonstrate that similar practices are common across the industry. In other words, Midjourney is trying to make a broader argument:
If the studios rely on comparable AI training methods themselves, can they fairly accuse others of doing the same?
Lawyers representing the studios have dismissed the request as a “fishing expedition,” arguing that Midjourney is attempting to obtain confidential information unrelated to the copyright claims. They also emphasize that the lawsuit is not intended to stop the development of AI. Instead, they say their goal is to prevent Midjourney from generating copyrighted characters without authorization.
The outcome of this lawsuit could shape the future of generative AI. Beyond determining Midjourney’s liability, the case may help define where fair use ends, what data AI models can legally be trained on, and how copyright law should apply to modern AI systems.
The ruling could ultimately become one of the most influential legal precedents for the entire AI industry, affecting not only Midjourney but also other leading AI companies developing generative models.
