Midjourney pushes Hollywood studios to disclose internal generative AI workflows
AI startup Midjourney is petitioning a court to compel Disney, Universal, and Warner Bros. to disclose internal documentation regarding their own use of generative AI technologies. The company argues that proving the studios engage in similar AI-driven workflows, including training on unlicensed content for storyboarding, is essential for its fair use defense in ongoing copyright litigation.
Key Takeaways
- Midjourney is challenging a June 2026 magistrate's ruling that limited discovery to "consumer-facing" AI applications.
- The startup seeks internal training datasets, model weights, and board presentations from major Hollywood plaintiffs.
- Studio attorney David Singer characterized the request as a "fishing expedition" designed to stall the infringement case.
- The legal demand includes all prompts and outputs generated by the studios on Midjourney's own platform.
Why It Matters
If Midjourney succeeds in broadening discovery, it could force a public accounting of how major studios use AI in their production pipelines before the content reaches audiences. For the streaming industry, this case tests the 'unclean hands' defense, which suggests rights holders cannot punish AI companies for the same training practices they utilize internally. A ruling in Midjourney’s favor would likely establish a precedent that internal R&D workflows are discoverable in copyright litigation. Watch for U.S. District Judge John A. Kronstadt’s decision on whether to uphold or overturn the current discovery limitations.
Additional Context
The pressure on Hollywood's AI strategy is mounting as studios move from experimentation to enterprise-scale integration. In September 2025, per Vice Chair Michael Burns, Lionsgate signed a first-of-its-kind deal with AI firm Runway to train a model on 20,000 film titles, specifically aimed at reducing costs in storyboarding and post-production. This move followed internal task forces established at Disney and Paramount in August 2025 to explore AI-driven cost savings and franchise expansion, though both companies have remained guarded about the specific datasets used for these projects. Simultaneously, the legal landscape for AI training has seen massive financial settlements. In September 2025, per Munck Wilson, Anthropic settled a major copyright case with authors for $1.5 billion, a figure that highlights the extreme liability risk for generative AI models trained on unlicensed works. This high-stakes environment is further complicated by shifting labor dynamics; the DGA reported a 35% drop in TV director employment in 2024, leading to stricter guild protections regarding the use of generative AI in production pipelines as established in July 2026. While studios like Disney have reportedly explored consumer-facing generative tools for Disney+—including features that would allow subscribers to create personalized content—the real friction lies in 'behind-the-camera' use. According to Deloitte's February 2025 analysis, while major studios are projected to spend less than 3% of their 2025 production budgets on generative tools, the rapid maturation of models for digital twins and voice cloning is outpacing current regulatory frameworks. The outcome of the Midjourney discovery dispute will likely determine if these internal efficiency gains remain trade secrets or become public evidence.
Read full article at techcrunch.com
Get this in your inbox → Subscribe
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source