Midjourney Fights Back as Hollywood Studios Seek to Keep Internal AI Strategies Secret in Copyright Battle

The Battle Lines of Hollywood versus Generative AI

The legal confrontation between major entertainment conglomerates and artificial intelligence developers reached a new peak after Disney, NBCUniversal, and DreamWorks initiated a copyright infringement lawsuit against the AI image-generation platform Midjourney. Filed in the U.S. District Court for the Central District of California, the complaint accused Midjourney of building its model by illegally training on vast libraries of copyrighted content. The studios, later joined by Warner Bros. Discovery, did not hold back in their legal filings, describing Midjourney as a quintessential copyright free-rider and a bottomless pit of plagiarism.

At the center of the studios' complaint is the allegation that Midjourney acts as a virtual vending machine, allowing users to generate near-identical replicas of highly protected characters with simple text prompts. The lawsuit highlighted side-by-side comparisons of official artwork and Midjourney-generated images of popular figures such as Darth Vader, Elsa from Frozen, Shrek, the Simpsons, Spider-Man, and the Minions. Pointing to Midjourney's reported revenue of over 300 million dollars, the plaintiffs are seeking damages and a preliminary injunction that could block the platform from operating without copyright safeguards, a move that could temporarily shut down the service.

Midjourney Counterattacks with Unclean Hands Defense

In response to the threat of a ruinous court ruling, Midjourney is mounting a defense centered on fair use and the legal doctrine of unclean hands. The image generator argues that the major Hollywood studios are hypocritical for suing over AI model training while simultaneously utilizing comparable AI methods within their own operations. By deploying the unclean hands defense, Midjourney aims to prove that the plaintiffs are engaging in the very same training practices they are attempting to punish in court.

Midjourney's strategy is designed to expose whether the film studios have secretly utilized AI models trained on unlicensed, third-party intellectual property to produce their own movies and television series. If the studios are found to be using unlicensed content for their internal AI programs, Midjourney argues that their legal standing and the credibility of their copyright claims will be severely undermined.

The High-Stakes Fight Over Secret Internal AI Records

The legal battle has now intensified into a dispute over how much internal information the studios must disclose. In June, Magistrate Judge Joel Richlin restricted Midjourney's discovery access, ruling that the studios only had to hand over information regarding their consumer-facing AI applications. Midjourney is actively fighting this limitation, with its attorney Bobby Ghajar filing a motion asking U.S. District Judge John A. Kronstadt to overturn the magistrate's ruling and grant access to the studios' private operational data.

Midjourney is seeking a wide array of highly confidential materials, including internal AI business plans, research reports, training datasets, model weights, board presentations discussing AI strategy, and documentation of how AI is integrated into film and television production. Ghajar argued in the filing that if the plaintiffs are executing the exact practices they want to penalize, that proof goes to the absolute heart of Midjourney's fair use and unclean hands defenses.

Should the court force Hollywood's biggest players to open their private books, the resulting disclosures could expose closely guarded industry secrets and redefine the boundaries of corporate AI adoption.