I. History
In June 2025, Disney, Universal, and their affiliated production companies, filed a federal lawsuit against Midjourney, a popular generative AI image platform used widely by commercial clients. This landmark lawsuit alleges that Midjourney has engaged in the “calculated and willful” copyright infringement of their intellectual property (Subin 2025). The studios also reference Midjourney's problematic training datasets, which scrape millions of copyrighted images from the internet without lawful permission. These training sets have allowed their 3.5 million monthly visitors to generate high-quality images that closely mimic or impersonate recognized fictional characters, such as Elsa, from Disney’s “Frozen” franchise, Darth Vader from Disney’s “Star Wars” franchise, and Minions, from Universal’s “Despicable Me” franchise. Hollywood studios allege that Midjourney’s business practices can be described as an “endless pit of plagiarism” (Subin 2025).
These accusations of piracy have caused Disney, Universal and their affiliate studios to seek an injunction to prevent Midjourney from continuing current practices that allegedly result in direct infringement through derivative works of copyrighted intellectual property. Hollywood studios have described Midjourney’s actions of garnering and profiting from their user base interactions, as a practice of secondary infringement, and demanded that this be halted as well. Disney and Universal are among the first corporations who have pursued litigation in order to defend their intellectual property against AI’s text-to-image generation capabilities.
In its defense, Midjourney argued that the studios “cannot stop AI training”. The company further claimed that AI training and their related training sets of aggregate data is protected under the Fair Use doctrine, stating that “training a generative AI model to understand concepts by extracting statistical information embedded in copyrighted works is a quintessentially transformative fair use” (Midjourney Inc. Answer to Complaint, 2025, lines 21-23). Midjourney’s lawyers contended that copyright law does not equate to having full control over the derivative works of copyrighted content. They also accused Disney and Universal of hypocrisy, noting that the studios themselves utilize AI services in their business production practices. Midjourney claimed that Disney, Universal, and their affiliated studios “cannot have it both ways, seeking to profit from industry-standard AI training practices on the one hand, while on the other hand accusing Midjourney of wrongdoing for the same.”
As of August 2025, this lawsuit remains unresolved, and no ruling has been issued.
II. Policy Background/Implications
Policymakers and courts have yet to establish clear standards on copyright and Fair Use practices in the context of AI training and generated works. This gap in regulation has created widespread uncertainty for AI service providers, their users, and creative professionals alike.
The Midjourney lawsuit outlines the growing tensions between generative AI services and current copyright law. This issue mainly revolves around three key complications.
First, extensive AI models require large datasets for training, many of which are compiled through publicly available, but often copyrighted, images and texts. The necessity of such datasets, coupled with the transparency questions and ethical issues surrounding the data collection, has raised concern.
Second, AI-generated outputs can closely resemble existing copyrighted works. Since AI models are trained to identify patterns from their training data and reproduce stylistic elements, they often produce derivative works extremely close to the copyrighted content. One such example can be found below, revealing striking visual similarities.

A screenshot from the Disney and Universal v. Midjourney lawsuit compares an AI-generated Midjourney image (left) with an image from Disney's "Frozen" (right). US District Court in Los Angeles. (Moses 2025)
Lastly, companies providing generative AI services can profit significantly from their outputs that may infringe on existing copyrights, raising liability questions. The current lack of a framework for compensating copyright holders of the original works calls into question the legitimacy and ethics of commercial AI practices.
The August 2025 animation film “Bad Guys 2” demonstrates how production companies are responding to this issue. The film includes an official disclaimer prohibiting the use or inclusion of their intellectual property in the training sets of AI models.
III. Policy Options
The emergence of generative AI services has revealed a significant gap in current copyright law and its ability to properly regulate new technologies. This gap has been highlighted in the ongoing lawsuit between Disney and Midjourney, a case that directly correlates to the popular commercial use of generative AI services in the creative industry. Tensions regarding the ethics and legality of current commercial practices, seen in service providers like Midjourney, raises valid concerns regarding the difference between Fair Usage and copyright infringement.
Clarify the Fair Use and Copyright Scope for AI Training and Datasets. The dissemination of well-defined legal definitions relating to Fair Use and its application to regulating AI Training datasets is necessary for clarifying the range at which AI developers are allowed, by law, to access otherwise copyrighted works for their training and improvement of AI services. Through this proposed solution, a clear criteria for transformative use in AI-generated works will assist both courts of law and creatives in understanding the clear-cut difference between allowed derivatives and infringement of original works.
Mandatory Transparency of Training Practices and the Introduction of Licensing Schemes. AI developers should be required to disclose all sources utilized and compiled in their training datasets. This act of mandatory transparency will ensure copyrighted works are not being infringed upon, protecting both the legal rights of creatives and the credibility of AI developers. A licensing model is another helpful method that can build credibility and trust, through promoting agreements where content creators can opt in to the usage of their works in training datasets. These schemes should be based on revenue-sharing models that benefit the copyright holders, protecting the intellectual property of artists and studios while enabling AI systems to continue learning responsibly.
Failure to comply with transparency or licensing standards should result in clear enforcement mechanisms and the immediate filtering and removal of infringing content must be carried out once notified.
IV. Conclusion
The lawsuit between Disney and Midjourney highlights the urgent need for updated copyright laws in the era of generative AI. The case raises important questions about how Fair Use should be applied and whether AI companies should be held accountable for outputs that closely resemble copyrighted works. Without updated legislation, those in the creative industry may feel unprotected while AI developers face uncertainty about what is legally allowed. Moving forward, lawmakers must prioritize adaptive approaches that balance AI innovation with ethical responsibility.
V. References
Daniel, George. 2025. “Disney & Universal Sue Midjourney: The AI Copyright Showdown That Could Change Hollywood Yoda in Court? Disney and Universal's Landmark AI Lawsuit Explained.” Lawyer Monthly. https://www.lawyer-monthly.com/2025/08/disney-universal-vs-midjourney-inside-the-ai-copyright-battle-that-could-rewrite-hollywood-law/.
Espiner, Tom. 2025. “Artificial intelligence: Disney and Universal sue Midjourney over copyright.” BBC. https://www.bbc.com/news/articles/cg5vjqdm1ypo.
Holz, David, and Justin McGill. 2025. “54+ Unique Midjourney Statistics - Users, Growth, Revenue.” BrandWell. https://brandwell.ai/blog/midjourney-statistics/.
Jentetics, Kinga. 2025. “Disney vs Midjourney: What This Landmark AI Lawsuit Means for Authors and Publishers.” Publish Drive. https://publishdrive.com/disney-vs-midjourney-what-this-landmark-ai-lawsuit-means-for-authors-and-publishers.html.
Kinsella, Eileen. 2025. “Midjourney Fires Back in Copyright Clash With Disney and Universal.” Artnet News. https://news.artnet.com/art-world/midjourney-fires-back-disney-and-universal-copyright-suit-2675798.
Knibbs, Kate, and Reece Rogers. 2025. “'Wall-E With a Gun': Midjourney Generates Videos of Disney Characters Amid Massive Copyright Lawsuit.” WIRED. https://www.wired.com/story/midjourney-generates-videos-of-disney-characters-amid-massive-copyright-lawsuit/.
Maddaus, Gene. 2025. “Midjourney Says Disney Cannot Prohibit AI Training.” Variety. https://variety.com/2025/film/news/midjourney-disney-ai-training-lawsuit-1236481167/.
Metzeling, Mark, and Jason Chan. 2025. “Disney and NBCUniversal Sue Midjourney in Landmark AI Copyright Case.” Macpherson Kelly. https://mk.com.au/disney-and-nbcuniversal-sue-midjourney-in-landmark-ai-copyright-case/.
Moses, Lucia. 2025. “Disney Says These Images Show How Midjourney Steals Its Characters.” Business Insider. https://www.businessinsider.com/disney-universal-midjourney-images-evidence-copyright-infringement-lawsuit-2025-6.
Penny, Brian, and Michaela Gordoni. 2025. “This Animated Comedy Includes Legal Warning to Ward Off AI Training.” Movieguide. https://www.movieguide.org/news-articles/the-bad-guys-2-includes-legal-warning-to-ward-off-ai-training.html.
Shibu, Sherin. 2025. “‘Bottomless Pit of Plagiarism’: Disney, Universal File the First Major Hollywood Lawsuit Against an AI Startup.” Entrepreneur. https://www.entrepreneur.com/business-news/disney-universal-sue-ai-startup-midjourney-plagiarism/493100.
Subin, Samantha. 2025. “Disney and Universal sue AI image company Midjourney for unlicensed use of Star Wars, The Simpsons and more.” CNBC. https://www.cnbc.com/2025/06/11/disney-universal-midjourney-ai-copyright.html.
“Transformative Use and Copyright Infringement Lawsuits | Intellectual Property Law Center.” 2024. Justia. https://www.justia.com/intellectual-property/copyright/fair-use/transformative-use/.