The legal battle between generative AI pioneer Midjourney and three titans of the entertainment industry—Disney, Universal, and Warner Bros.—has entered a contentious new phase. What began as a traditional copyright infringement lawsuit has transformed into a high-stakes discovery dispute that threatens to pull back the curtain on how Hollywood’s own creative giants are experimenting with the very technology they seek to penalize in the courtroom.
Midjourney, currently embroiled in litigation regarding its image-generation models’ ability to replicate iconic characters like Darth Vader and Batman, is now aggressively moving to compel the studios to disclose their own internal AI development practices. The startup argues that if the studios are utilizing generative AI for storyboarding or concept development, it exposes a fundamental hypocrisy in their legal claims, suggesting that the industry’s own standards may mirror the "infringing" behaviors they accuse Midjourney of perpetrating.
The Chronology of the Conflict
The legal collision course between Silicon Valley and Hollywood began in earnest last year when Disney and Universal filed a joint lawsuit against Midjourney. The plaintiffs alleged that the startup’s text-to-image models were trained on unauthorized datasets containing their intellectual property, resulting in the ability to generate hyper-realistic depictions of characters protected by copyright law.
Months later, Warner Bros. joined the fray, filing its own complaint against Midjourney. The core of the studios’ argument is that Midjourney’s model architecture is built upon a foundation of stolen creative assets. By feeding their films and series into the training set, the studios contend, Midjourney has created a machine that directly competes with the original creators by commodifying their most valuable character brands.
Midjourney has consistently maintained a stance of "fair use," arguing that the transformative nature of its technology falls under the protections of copyright law. As the discovery phase—the portion of a lawsuit where both sides share evidence—has progressed, the focus has shifted from the final images generated by Midjourney to the internal processes of both the startup and the studios.
In a previous ruling, a judge ordered that the studios must provide documentation regarding their use of generative AI, but with a significant caveat: the disclosure requirement was limited to AI usage that resulted in "consumer-facing" content. Midjourney is now seeking to overturn this limitation, arguing that it prevents them from uncovering the "behind-closed-doors" truth about the studios’ own AI R&D.
The "Fair Use" Defense and the Discovery Tug-of-War
Midjourney’s latest court filing is a direct challenge to the current parameters of discovery. The startup argues that by limiting the scope of document production to consumer-facing AI products, the court is inadvertently allowing the studios to engage in a form of strategic information hoarding.
"The studios are cherry-picking only those documents they believe support their market harm claims while depriving Midjourney of documents that would support its defenses," the filing states.
The "Hypocrisy" Argument
Midjourney’s legal team is pushing for a broader interpretation of discovery, asserting that if the studios are developing internal AI models for storyboarding, character design, or pre-production concept art, that evidence is legally material. The implication is clear: if Hollywood studios are training their own AI models on copyrighted material for internal efficiency, it establishes an "industry custom."
If Midjourney can prove that the studios—the very entities claiming that training AI on copyrighted data is a violation—are doing exactly that within their own walls, it could significantly weaken the studios’ case. It shifts the narrative from "Midjourney is a thief" to "Midjourney is adopting an industry-standard practice that the studios only dislike when they aren’t the ones doing it."
The Prompting Dispute
Beyond internal development, Midjourney is also demanding that the studios turn over a comprehensive history of the prompts used to interact with Midjourney’s models. The studios want the discovery process to focus solely on the prompts that generated allegedly infringing content—essentially, the prompts that resulted in characters like Bart Simpson or Superman.
Midjourney, however, argues that this is an incomplete picture. They are seeking all prompts used by the studios, regardless of whether the output was flagged as infringing. This, they argue, is necessary to establish the context of how these models are being used and to test the limits of the studios’ own engagement with the technology.
Official Responses and Legal Posturing
The studios’ legal representatives have reacted sharply to Midjourney’s aggressive discovery tactics. Lead attorney David Singer has characterized Midjourney’s requests as a "fishing expedition."
According to Singer, the studios are not attempting to stifle technological progress or force Midjourney out of business. "The studios do not seek to stop AI technology or even shut down Midjourney’s business," Singer stated in a previous hearing. Instead, he framed the lawsuit as a measured attempt to protect intellectual property rights. The studios maintain that they simply want Midjourney to cease the unauthorized reproduction, distribution, and public display of their copyrighted characters.
From the studios’ perspective, the internal use of AI—whether for storyboarding or ideation—is irrelevant to the charge that Midjourney has built a commercial enterprise on the back of stolen data. They argue that Midjourney is attempting to distract the court from the central issue: that the startup’s business model is fundamentally predicated on the misappropriation of creative works without permission or compensation.
Broader Implications for the Creative Economy
This lawsuit represents a critical juncture for both the entertainment industry and the AI sector. The implications of the final ruling will likely set a precedent for how intellectual property is managed in the age of generative intelligence.
The Standard of Industry Practice
If the court rules in favor of Midjourney and forces the studios to open their internal development logs, it could set a new standard for corporate transparency. The entertainment industry has historically been protective of its internal "secret sauce." If it is revealed that studios have been utilizing AI in ways that parallel the models they are suing, it could lead to a massive cultural and legal shift in how AI training is regulated.
The Future of Creative Labor
For Hollywood’s creative workforce—writers, illustrators, and concept artists—this case is a harbinger of potential disruption. If studios are indeed using AI to storyboard and develop characters, it raises questions about the future of creative labor. Are these tools being used to empower artists, or are they being used to replace the human element of the creative process?
The Legal Definition of "Transformative"
Ultimately, the outcome of this dispute will hinge on how the courts define "fair use" in the context of large-scale machine learning. If training an AI model is deemed "transformative" enough to bypass copyright restrictions, it could pave the way for a new era of generative tools. If not, companies like Midjourney may face a future of forced licensing, where they must negotiate with the very studios currently trying to dismantle them.
As the discovery phase continues, the tension between these two industries remains at a fever pitch. Whether this case leads to a landmark settlement or a protracted trial that reaches the Supreme Court, one thing is certain: the era of "black box" AI development is ending, and the era of legal accountability is just beginning. Both sides are digging in, and for now, the courtroom has become the primary arena where the future of human-AI collaboration will be decided.
Disclaimer: This article is intended for informational purposes and does not constitute legal advice. When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Reynand Wu
Lina Irawan
