Midjourney’s Counterattack: Making Hollywood’s AI Secrets the Issue
Midjourney’s legal counterstrategy in the current Midjourney copyright lawsuit is to force major studios to reveal how they train and use artificial intelligence, turning a narrow infringement case into a broader fight over Hollywood AI training and internal generative tools.
Instead of treating the lawsuit from Disney, Universal and Warner Bros. Discovery as a one-way attack, Midjourney is trying to put the studios’ own practices on the record. The studios sued last year, accusing the AI image generator of copyright infringement for being able to generate images of Superman, Batman and other copyrighted characters. The case began in June 2025, when Disney and Comcast’s Universal accused Midjourney of enabling large-scale infringement involving characters like Darth Vader and Elsa. Warner Bros. Discovery later joined, calling Midjourney’s conduct “brazen theft” and seeking USD 150,000 (approx. RM690,000) per infringed work. Midjourney’s response is blunt: if it falls, the studios should be dragged into the same spotlight.
The Discovery Gambit: Forcing Studios to Open the AI Black Box
Midjourney’s most aggressive move so far lies in AI copyright discovery. The company is asking a federal judge to compel Disney, Universal and Warner Bros. Discovery to disclose their internal use of artificial intelligence. It wants to see how they use AI technologies in their shows and movies and have that information submitted to the court.
This is not a modest request. Midjourney is seeking the studios’ AI business plans, research reports, training datasets, model weights and even AI presentations prepared for their boards. In other words, it wants the roadmap, the fuel and the engine of Hollywood’s own AI systems. A magistrate judge has already limited this, ruling that the studios only need to provide information on “consumer-facing” AI tools rather than internal systems. Midjourney is now urging the federal court and Judge John Kronstadt to overturn that restriction, arguing that the real story is hidden behind internal firewalls.
Fair Use, Unclean Hands, and the Charge of Hypocrisy
Midjourney’s legal theory is simple and confrontational: training AI with publicly available images is fair use, and major studios allegedly rely on the same methods for their own models. More than a year into the litigation, the company has leaned on a fair use defense and claims the studios engage in comparable AI practices internally.
The central accusation is not subtle. Midjourney’s attorney, Bobby Ghajar, argues that if the plaintiffs are “doing the very thing they seek to punish,” that evidence is vital to both fair use and an “unclean hands” defense. In plain language, Midjourney is calling out what it sees as Hollywood’s double standard: condemning AI startups for training on copyrighted content while quietly building their own generative tools the same way. The studios’ lawyer has dismissed this as a “fishing expedition” meant to distract from Midjourney’s conduct. But even if the court partially agrees, the moral narrative has already shifted: the accusers may need to defend their own AI ethics.
What This Battle Means for AI, Studios, and Future Lawsuits
Midjourney’s push is not only about survival in one Disney AI legal battle; it is about shaping the rules of AI copyright discovery for everyone. If the federal court decides that internal AI plans, datasets and model weights are fair game in litigation, future plaintiffs may have to expose their own systems before attacking others.
That prospect should worry legacy media companies as much as AI startups. The decision on whether to overturn the magistrate’s narrow discovery order will signal how transparent companies must be when they claim infringement in an AI context. Either outcome will carry precedent weight: allowing broad discovery could deter aggressive lawsuits; restricting it could encourage more selective, one-sided cases. Midjourney’s bet is that the studios will blink first rather than reveal their AI playbooks. If it is right, the headline lesson is uncomfortable for Hollywood: you cannot demand a clean AI ecosystem while hiding how your own models are trained.
