Hollywood’s AI Lawsuit Exposes an Industrial Theft Machine -By Fransiscus Nanga Roka

Artificial intelligence did not invent Darth Vader, Elsa, Spider-Man, Shrek or the Minions. Human creators did. Yet Midjourney allegedly built a lucrative image-generation business by absorbing copyrighted works and enabling users to manufacture remarkably faithful replicas within seconds. That is not merely technological disruption. It is a test of whether industrial-scale copying becomes lawful when hidden inside an algorithm.

On June 11, 2025, Disney Enterprises, Marvel, Lucasfilm, 20th Century Fox, Universal and DreamWorks sued Midjourney in the US District Court for the Central District of California—the first direct copyright action by major Hollywood studios against a generative-AI company. The complaint portrays Midjourney as a “bottomless pit of plagiarism” and alleges that it copied protected works for training while distributing unauthorized images of famous characters. Reuters and the Associated Press documented the lawsuit’s central claims.

The dispute’s who, what, when and where are straightforward. Its why and how are explosive.

Why did the studios sue? Because Midjourney allegedly continued generating infringing material despite prior demands for protective measures. How does the alleged infringement occur? First, copyrighted images are purportedly copied into training pipelines. Second, prompts can generate outputs resembling protected characters closely enough to compete with licensed illustrations, merchandise and entertainment content.

Midjourney’s anticipated shield is fair use: training, it argues, transforms publicly available images into a system capable of producing new expression. But “transformative” cannot function as a magic word that legalizes every commercial appropriation. Fair use requires a contextual assessment of purpose, nature, amount copied and market harm. A model that reproduces recognizable protected expression on demand is materially different from one that merely learns abstract visual concepts.

The 2025 ruling against Ross Intelligence is instructive. A federal court rejected fair use where copyrighted Westlaw material was copied to create a competing legal-research product. It does not decide Midjourney’s case, but it demolishes the fantasy that AI training is automatically immunized from copyright law. Reuters

Midjourney may scrutinize whether studios themselves use generative AI or unlicensed third-party material. Such discovery could expose Hollywood’s hypocrisy—and it should. But “unclean hands” is not a universal permission slip: Midjourney must establish misconduct sufficiently connected to the rights the studios seek to enforce. One actor’s wrongdoing does not automatically legalize another’s.

Hollywood is hardly an innocent guardian of struggling artists. Studios have historically concentrated copyrights, weakened bargaining power and monetized creativity long after individual creators lost control. Nevertheless, corporate hypocrisy cannot erase valid copyrights. The court must protect creators without converting dominant studios into permanent tollbooths over culture.

A strategically sound judgment should require four things: auditable disclosure of training-data provenance; effective notice-and-removal and character-level safeguards; compensation or collective licensing when protected works materially train commercial models; and proportionate liability based on knowledge, output similarity and market substitution.

Congress should establish mandatory training-data transparency, independent audits and a statutory licensing mechanism accessible to small creators—not merely conglomerates. Courts should also reject both extremes: AI exceptionalism, which launders mass copying as innovation, and maximalist copyright enforcement, which allows powerful owners to monopolize styles, ideas and cultural reference.

The real question is not whether AI should survive. It is whether AI companies may privatize humanity’s creative heritage, conceal the copying, sell the replicas—and call the entire operation progress.

Fransiscus Nanga Roka

Faculty of Law University 17 August 1945 Surabaya and Managing Partner Law Firm Victorious Indonesia