Midjourney wants Disney, Universal, and Warner Bros. to reveal in court how they use AI across their companies; studios sued Midjourney in 2025 for infringement
Context & Ripple Effects
Disney and NBCUniversal’s initial California case framed Midjourney as both a training-data and user-output infringement risk after talks reportedly failed. Warner Bros. Discovery later brought a parallel claim centered on its characters, widening the studio front against the image generator.
The discovery request shifts attention from Midjourney’s conduct alone to the plaintiffs’ own AI practices. That matters because Disney has separately pursued controlled AI arrangements, indicating the dispute is over unauthorized use and governance of IP rather than a blanket rejection of generative AI.
First-order effects
- Disney, Universal, and Warner Bros. may have to produce internal information about AI deployment, IP controls, and related policies if the court grants Midjourney’s request.
- The studios’ infringement claims could face closer scrutiny over how their own AI use and licensing posture bears on requested remedies or factual arguments.
Second-order effects
- Discovery could sharpen the commercial distinction between AI vendors operating under studio oversight and tools that studios allege were trained on or generate protected characters without permission.
- Other image and video AI providers facing studio claims, including MiniMax, will have added incentive to document content provenance, rights controls, and safeguards around recognizable franchise material.
Third-order effects
- If courts allow broad discovery into rightsholders’ internal AI use, copyright litigation may increasingly become a venue for defining what counts as controlled, licensable AI use of entertainment IP.
- The broader industry is likely to sort between negotiated, governed access to valuable media catalogs and litigation over unlicensed training and outputs; the eventual boundary remains dependent on court rulings and deal terms.
The trend: Generative-AI copyright fights are evolving from claims about model outputs into a contest over the licensing, controls, and evidence that govern commercial use of premium IP.