Midjourney’s Big Bet: Turn Copyright Claims Back on Hollywood
Midjourney’s legal strategy is to use the discovery process in a copyright lawsuit to force major studios to reveal their own Hollywood AI training methods, arguing that the same practices they condemn in generative tools underpin internal studio AI systems and should be examined for double standards and fair use implications. This is not a side issue in the Midjourney copyright lawsuit; it is the core of the company’s defense. Disney and Universal sued in June 2025, accusing Midjourney of enabling large-scale infringement of copyrighted characters such as Darth Vader and Elsa. Warner Bros. Discovery followed three months later, calling Midjourney a “brazen theft” engine and seeking USD 150,000 (approx. RM705,000) per infringed work. Midjourney’s reply is blunt: if studios train their own models on copyrighted works, their moral outrage looks less like principle and more like a business tactic.
What Midjourney Wants to See Inside Studio AI Labs
Midjourney is not asking for a marketing brochure; it wants the plumbing of studio AI systems exposed. The company has requested AI business plans, research reports, training datasets, model weights, and board presentations on AI from Disney, Universal, and Warner Bros. Discovery. That goes far beyond consumer tools and into how Hollywood AI training methods are designed and justified. A magistrate judge has tried to narrow this generative AI legal discovery, allowing studios to withhold most internal information and disclose only details tied to consumer-facing AI applications. Midjourney has now asked Judge John Kronstadt to overturn that decision, arguing the restricted record makes its fair use defense impossible to test against real studio behavior. This is a calculated pressure move: the company is betting that studios want to keep their own AI practices disclosure minimal because those practices look a lot like the ones they are suing.
Fair Use vs. Unclean Hands: Calling Out Hypocrisy
Midjourney’s argument is not subtle: training on publicly available images is fair use, and the studios do the same with their internal models. Its attorney Bobby Ghajar wrote that if the plaintiffs are “doing the very thing they seek to punish,” that evidence goes to the heart of Midjourney’s fair use and unclean hands defenses. In plain terms, Midjourney says Hollywood’s AI adoption is hypocritical. Studios want the creative and financial upside of generative systems while treating independent AI firms as a “bottomless pit of plagiarism” when those firms lack studio-sized licensing clout. That double standard matters. If discovery shows studios quietly training models on copyrighted works in ways that look similar to Midjourney’s approach, then the dispute stops being about principle and starts being about who is allowed to automate creativity. This is why the fight over generative AI legal discovery is so bitter: it threatens to collapse the distinction between the accusers and the accused.
Why This Discovery Fight Could Reshape AI Copyright Law
The outcome of this discovery skirmish could set the first real precedent on how courts evaluate AI training practices in copyright disputes. If Judge Kronstadt expands the scope of studio AI practices disclosure, future plaintiffs may have to expose their own generative tools before claiming infringement by others. That would discourage opportunistic lawsuits built on PR outrage rather than clean hands. Conversely, if the narrow ruling stands and only consumer-facing AI use is visible, large rights-holders can continue policing external AI while hiding their internal experiments. Either way, Midjourney has already changed the tone of the Midjourney copyright lawsuit: studios can no longer pretend to be AI bystanders. They are active participants whose Hollywood AI training methods are now part of the story. The court’s next move will decide whether that story is told in full or kept behind closed doors.
