Midjourney Seeks Disney, Universal, and Warner Bros. AI Records in Copyright Fight
Midjourney has asked a federal judge to compel Disney, Universal, and Warner Bros. to disclose how they use artificial intelligence, escalating a copyright dispute that began with the studios’ claims against the image generator. The company wants Judge John Kronstadt to overturn a June 15 ruling that denied broad discovery into the studios’ internal AI tools.
At the center of the motion is Midjourney’s argument that the studios may be using the same technology they accuse it of exploiting. In filings reported by Variety, Midjourney attorney Bobby Ghajar said that if the plaintiffs are doing “the very thing they seek to punish,” that evidence would matter to the company’s fair use and unclean hands defenses. Midjourney is seeking information on business plans, training datasets, model weights, and even board presentation materials.
The studios sued Midjourney over allegations that it facilitates the unauthorized reproduction of copyrighted characters, including Superman, Scooby-Doo, and Bugs Bunny. Their attorney, David Singer, dismissed the new request as a “fishing expedition” meant to divert attention from Midjourney’s legal exposure. He said the studios want the company to stop copying their movies and television shows and to halt the creation and distribution of works containing unauthorized versions of their characters.
The dispute adds another layer to a company that has faced copyright scrutiny since its public launch. In 2023, three artists filed a class-action complaint alleging that Midjourney had been trained on billions of scraped images and could generate works in artists’ styles without consent or compensation. That case, filed in the Northern District of California against Midjourney, Stability AI, and DeviantArt, is still moving forward, with some claims dismissed and others allowed to proceed.
Midjourney drew renewed criticism in 2024 after artists circulated a 24-page Google Sheet naming 16,000 artists whose work they said had been used as training data. The list included Pablo Picasso, Frida Kahlo, Andy Warhol, and Yayoi Kusama, along with the work of a six-year-old child who had participated in a 2021 fundraiser for Seattle Children’s Hospital. The material was later folded into an amended class-action complaint filed in November 2023.
For now, the case is becoming less about one company’s tools than about a broader question now moving through the courts: whether the AI industry can challenge copyright claims while keeping its own training practices out of view.
