Warner Bros sues Midjourney over AI images of Superman, Batman, Bugs Bunny

Close-up editorial portrait of David Holz against a split backdrop of glowing courthouse columns and abstract AI diffusion patterns, warm amber on one side and cool electric blue on the other, a judge's gavel in soft focus foreground with faint generic silhouettes of a caped superhero and cartoon rabbit ears, clean high-key lighting, crisp details, shallow depth of field, without any text or logos

Warner Bros. has filed a copyright infringement lawsuit against Midjourney, alleging the AI company enables users to generate images and videos of characters such as Superman and Bugs Bunny. According to Yahoo Entertainment, the complaint argues that Midjourney trained on “illegal copies” of Warner Bros. works and encourages prompts invoking well-known characters.

Allegations detail training data and user prompts

The lawsuit, filed in Los Angeles federal court, states that Midjourney’s system was trained on unauthorized copies of Warner Bros. material and that the service promotes selecting characters like Batman, Wonder Woman, Scooby-Doo, and the Powerpuff Girls to create downloadable images and videos in “every imaginable scene.” It further claims that even a generic prompt for a “classic comic book superhero battle” produces high-quality images of DC figures such as Superman, Batman, and Flash.

Warner Bros. asserts that “Midjourney thinks it is above the law” and could restrict the use of its service to prevent infringement, noting that the platform already sets limits on content such as violence or nudity. The filing alleges that Midjourney’s practices foster “consumer confusion regarding what is lawful and what is not lawful” by implying authorization where there is none, and seeks up to $150,000 in damages per infringed work.

Context within broader Hollywood litigation

The case marks the third major studio action against Midjourney in the same court after Disney and Universal filed a joint lawsuit in June. Midjourney, based in San Francisco, did not immediately respond to a request for comment, Yahoo reports.

In an August filing responding to the Disney and Universal case, Midjourney denied infringement and said its AI “had to be trained on billions of publicly available images” to learn visual concepts and their correspondence to language. The company argued that training a generative AI model by extracting statistical information from copyrighted works is “quintessentially transformative fair use,” referencing court rulings cited in that response. Midjourney also said users are responsible for following its terms of use, which prohibit infringing intellectual property rights.

In a 2022 interview with The Associated Press, CEO David Holz compared the service to a search engine pulling images from across the internet and likened AI learning to how people learn from existing works. Holz suggested that if resulting images “come out differently then it seems like it’s fine.”

Total
0
Shares
Pridaj komentár

Vaša e-mailová adresa nebude zverejnená. Vyžadované polia sú označené *

Previous Post
Close-up editorial portrait of Jensen Huang in a black leather jacket against a luminous data center corridor, warm rim light on his profile and cool blue server glow behind, foreground shows three unbranded silicon chips hovering as clean geometric forms with one radiant gold chip sharp in focus and two cooler-toned chips receding, high-key brightness, crisp detail, no text or logos, neutral expression.

Broadcom Soars on Reported $10B OpenAI Chip Order; Nvidia, AMD Slip

Next Post
A luminous golden key formed from delicate circuit traces turning a sleek door lock that opens into a bright workspace filled with diverse human silhouettes, warm gold light spilling through into a cool blue corridor, close medium framing with bold shapes and high-key contrast.

Salesforce, Klarna cut roles as OpenAI builds jobs platform to reskill millions

Related Posts