Midjourney has responded to a lawsuit filed by Disney and Universal in June, asserting that the studios lack the authority to prevent artificial intelligence (AI) training on their copyrighted works. The studios had accused Midjourney, an AI image platform startup, of “vast, intentional, and unrelenting copyright infringement,” alleging that platform users could generate nearly identical reproductions of the studios’ proprietary characters.
In its response, filed on Wednesday evening (via Variety), Midjourney contended that AI training qualifies as “fair use.” Midjourney’s legal counsel argued that “Copyright law does not confer absolute control over the use of copyrighted works.” They further elaborated that “The limited monopoly granted by copyright must give way to fair use, which safeguards countervailing public interests in the free flow of ideas and information.”
Midjourney also asserted that the studios are attempting to “have it both ways” by utilizing AI tools themselves while simultaneously pursuing legal action against a prominent AI service. The filing indicated that Midjourney is a widely used tool among visual effects companies and other vendors that collaborate with Disney and Universal. The response additionally stated that “many dozens” of Midjourney’s subscribers possess email addresses linked to Disney and Universal, suggesting that employees of the studios are also engaging with the service.
Midjourney’s lawyers noted that Disney CEO Bob Iger expressed approval of AI during an annual meeting in March, stating that “technology is an invaluable tool for artists.” The filing explicitly stated, “Plaintiffs cannot have it both ways, seeking to profit — through their use of Midjourney and other generative AI tools — from industry-standard AI training practices on the one hand, while on the other hand accusing Midjourney of wrongdoing for the same.”
Midjourney rolls out V1 AI video generation model
The lawsuit initiated by Disney and Universal primarily focuses on AI outputs, maintaining that Midjourney users are creating images substantially similar to the studios’ copyrighted works. This distinguishes the suit from other legal challenges that have solely argued that AI training itself constitutes infringement. Midjourney’s defense highlights that its users are obligated to comply with the platform’s terms of service, which prohibit the infringement of intellectual property rights. However, the company’s lawyers argued that merely producing images similar to copyrighted works is insufficient to demonstrate infringement.
Midjourney’s legal representation stated, “Indeed, there are any number of legitimate, noninfringing grounds to create images incorporating characters from popular culture like those claimed by Plaintiffs, including non-commercial fan art, experimentation and ideation, and social commentary and criticism.” They concluded that “Plaintiffs seek to stifle them all.” Midjourney is represented in this legal matter by Bobby Ghajar, John Paul Oleksiuk, Judd Lauter, and Ellie Dupler of Cooley LLP. Mr. Ghajar also represents Meta in a separate case where authors have accused that company of illicitly training its AI language model using their published books.
