Warner Bros. Discovery Copyright Lawsuit Targets Midjourney for Generating Batman and Superman Images

Not even Batman and Superman can escape artificial intelligence (AI). On Thursday, Warner Bros. Discovery filed a lawsuit against Midjourney for allegedly reproducing, displaying, and distributing unauthorised derivatives of its intellectual property. The media and entertainment company shared dozens of images of its copyrighted work, including characters such as Batman, Superman, Scooby Doo, Rick and Morty, Bugs Bunny, and more. The company is seeking monetary compensation for damages and wants the courts to block Midjourney from generating copies of its work.

Midjourney Faces Lawsuit for Generating AI Images and Videos of Batman

Warner Bros. Discovery filed the lawsuit with a US District Court in California for direct and secondary copyright infringement of its intellectual property, seeking a jury trial. The company alleged that Midjourney “brazenly dispenses” its copyrighted works “as if it were its own.” Across the entire lawsuit, numerous examples of AI images generated by Midjourney were presented.

Image taken from the Warner Bros. Discovery v Midjourney lawsuit
Photo Credit: Warner Bros. Discovery

The media and entertainment company claimed that these listed examples were obtained by providing a text prompt asking to see these characters in a specific pose or location. Warner Bros. Discovery also claimed that some of the outputs with striking resemblance to its characters were generated without even mentioning them. For instance, the lawsuit claims, the prompt, “classic comic book superhero battle,” generated an image featuring Superman and Batman.

As per the lawsuit, several examples shared were taken from Midjourney's Discord channel and Subreddit. Highlighting this, the company said it constitutes proof that the company knew of its service's infringement of the copyrighted characters.

Image taken from the Warner Bros. Discovery v Midjourney lawsuit
Photo Credit: Warner Bros. Discovery

Warner Bros. Discovery alleges that Midjourney could have easily stopped its AI image generation and video generation models from infringing on its copyright; however, “Midjourney has made a calculated and profit-driven decision to offer zero protection for copyright owners.” The lawsuit claims that the AI company had full knowledge of the scope of its AI systems' piracy.

Further, the lawsuit also raised concerns that such immaculate reproduction would not have been possible unless Midjourney's large language models (LLMs) were trained on its copyrighted data. It claimed that the AI firm's CEO, David Holz, trained its model on copyrighted images by sourcing them from the Internet. To verify the claims, the entertainment company stated that the training phase of Midjourney's AI models will be part of the discovery of this legal procedure.