On June 11 (U.S. time), Hollywood film studios Disney and Universal Pictures—along with their subsidiaries Pixar, Marvel, DreamWorks, and others—jointly filed a lawsuit against AI image generation company Midjourney, accusing it of enabling users to generate unauthorized depictions of virtually all characters from their animated films, including the Minions, Frozen, and Shrek. They characterized the company’s behavior as a blatant case of “free-riding and unrestrained plagiarism.”
This lawsuit marks the first time Hollywood has taken legal action to defend its intellectual property (IP) rights in response to the rise of AI-generated multimodal content like images and videos. Regardless of the case’s outcome, it is poised to influence the commercial models of AI companies that rely on copyrighted content for training and generation.
Midjourney, a U.S.-based startup, publicly launched its text-to-image tool of the same name in July 2022 and quickly amassed a large user base. As of now, Midjourney’s Discord server hosts more than 21 million users. The company operates under a SaaS subscription model and, according to its founder David Holz, had already become profitable as early as August 2022. In April 2024, Midjourney released its latest version, Midjourney V7.
In the lawsuit, Disney and Universal Pictures (hereinafter “the plaintiffs”) allege that Midjourney earned $300 million in revenue in 2024. They provided specific examples demonstrating how the platform enables users to seamlessly generate infringing content. When prompted with terms like “Darth Vader,” “Minions,” or “Shrek,” Midjourney was capable of producing high-quality images nearly identical to the original animated characters.
The plaintiffs argue that Midjourney scraped data related to copyrighted works from the internet using various tools. The AI-processed content was then used to generate results for users. Through reverse image tracing, the plaintiffs found that several outputs bore strong resemblances to scenes from the original films.
The lawsuit cites Holz’s own words from an August 2022 interview, in which he admitted that Midjourney collected “all data, all texts, all images” for training, storing them in a large dataset, which was then “ignited” to produce something massive. In a September 2022 interview with Forbes, Holz also acknowledged that Midjourney had never sought permission to use any copyrighted material.
What Midjourney threatens, the plaintiffs assert, is one of Hollywood’s most vital revenue streams: IP licensing. Through official licensing agreements, character likenesses from Disney and Universal’s animated films can appear on clothing, stationery, and plush toys—with the studios collecting fees from manufacturers and brand partners. The business model of “licensing revenue exceeding box office, and international box office outperforming domestic” has long been a pillar of Hollywood’s global content dominance.
The plaintiffs liken Midjourney to a “vending machine” capable of endlessly generating infringing content—reproducing copyrighted characters such as Spider-Man, the Minions, and Iron Man, allowing users to freely download and share the images. Through its public “Explore” page, Midjourney showcases AI-generated works tied to these characters, making them easy for creators to find and reference, thereby attracting more users through high visual fidelity.
According to the plaintiffs, they had explicitly demanded that Midjourney cease its infringing activities prior to filing the lawsuit, but were ignored. They further argue that since Midjourney already restricts the generation of violent or sexually explicit content, it is fully capable of preventing copyright infringement as well. The plaintiffs also claim that Midjourney is preparing to launch video-related services without obtaining the necessary rights.
The lawsuit was filed in the U.S. District Court for the Central District of California, seeking an injunction against Midjourney to halt willful infringement and demanding compensation of up to $150,000 per infringed work.
As AI model developers scale by leveraging publicly available data, they inevitably face backlash from content rights holders.
A similar lawsuit was filed on June 24, 2024, when the world’s three major music companies—Sony Music Group, Universal Music Group, and Warner Music—along with several other record labels, jointly sued two AI music companies, Suno and Udio, accusing them of illegally using copyrighted music to train their models and provide services. That case remains unresolved.