Disney, Universal, and several other American film studios have filed a lawsuit against artificial intelligence company Midjourney, accusing its popular image generation tool of systematically infringing copyrights by creating unauthorized replicas of famous characters.
The lawsuit was submitted to the U.S. federal court on Wednesday, describing Midjourney as a "bottomless pit of plagiarism."
"By improperly using the plaintiffs' copyrighted works and then distributing images blatantly incorporating and copying Disney and Universal's iconic characters (soon to include videos) without spending a penny on their creation, Midjourney is a quintessential copyright freeloader," the complaint stated. "Piracy is piracy, whether the infringing images or videos are made with AI or other technologies, it does not diminish the nature of the infringement."
The studios, including DreamWorks and Twentieth Century Fox Film Corporation, cited examples of Midjourney’s outputs featuring Yoda, Marvel superheroes, characters from 'Aladdin,' Minions, 'The Simpsons,' and 'Shrek.'
They are seeking damages and an injunction to prevent the platform from copying, displaying, or distributing their copyrighted content. Decrypt has contacted Midjourney for comments on the lawsuit.
This case is part of a wave of lawsuits against AI companies for copyright infringement.
As generative AI tools become more prevalent, legal scrutiny over their training data and output is intensifying. The New York Times sued OpenAI and Microsoft in December 2023, while Reddit is currently taking legal action against Anthropic. Other plaintiffs include music publishers and media companies like Ziff Davis.
The core legal issue in these lawsuits is whether AI companies can legally use unlicensed copyrighted works during training or generation.
Companies like OpenAI have admitted to using copyrighted content in training and argued that developing AI systems without such content would be "impossible."
Filtering Issue
While some companies have sought licensing agreements with copyright holders, others like Midjourney have faced criticism for inadequate protective measures.
In the complaint, the studios argued that Midjourney has the capability to filter prompts and outputs, similar to how it currently blocks certain violent or pornographic content. They claim the company could easily implement comparable copyright protections but chooses not to.
Midjourney previously faced controversy for restricting the generation of images related to China’s President Xi Jinping while allowing the creation of images of other world leaders.
Meanwhile, the film industry is grappling with how AI will reshape creative labor. A report by the British Film Institute this week warned that AI poses a direct threat to jobs and revenue in the screen industry.
The research cited in the report suggests that global audiovisual creators' income could drop by up to 21% within three years, and by 2026, more than 200,000 entertainment jobs in the United States—especially entry-level positions—could be affected.