Home Artificial Intelligence Database of Artists Used to Train AI Leaks to the Public

Database of Artists Used to Train AI Leaks to the Public

A database of artistic styles allegedly used to train Midjourney’s generative artificial intelligence (AI) program is garnering criticism on social media after it began circulating online at the end of last year. The “Midjourney Style List” spreadsheet contains more than 16,000 artists’ names, including those of prominent living figures such as Banksy, David Hockney, Yayoi Kusama, KAWS, and even a six-year-old who created a drawing in 2021 for a hospital fundraiser. The list also specified time periods and artistic movements, mediums, genres, and video game softwares. 

While the spreadsheet has seemingly been made inaccessible, it is still viewable via the Internet Archive, and a court document filed in late November 2023. Containing a portion of the artists’ names listed in the database, the report is part of an ongoing class action lawsuit against DeviantArt, Midjourney, Stability AI, and Runway AI. Initially filed last January by 10 visual artists, the suit accuses the companies of copyright infringement for allegedly using their work without their permission to train AI. The development of unregulated generative AI has also led to other lawsuits, strikes, and congressional hearings over fears of future job loss and allegations of copyright violations.

Midjourney has not yet responded to Hyperallergic’s request for comment.

Court documents detail that the list was first published to a Discord server in February 2022 by Midjourney CEO David Holz, who is quoted in a series of messages explaining how the names can be used as text prompts to generate images. 

“I think you’re all gonna get mind blown by this style feature,” Discord messages attributed to Holz read. “I should be clear it’s not just genres. It’s also artist names. It’s mostly artist names. 4000 artist names, 1000 styles.”

On December 31, Riot Games artist Jon Lam posted screenshots of this chat to X, showing Holz and other Midjourney developers discussing how the database contained the work of at least 4,700 individual artists. According to these messages, the artists’ names were collected from Wikipedia and Magic the Gathering, an online trading card game. Additional screenshots posted by Lam further show messages attributed to software developer Brad Templeton warning Holz of potentially running into a “copyright problem” with training systems on real artists’ work.

Other high-profile artists named in the list include Dan Flavin, Salvador Dalí, David Lynch, Anish Kapoor, Andy Warhol, Claude Monet, Cy Twombly, Damien Hirst, Frida Kahlo, the Guerrilla Girls, Joan Mitchell, Joan Miró, Marcel Duchamp, Richard Serra, Jean-Michel Basquiat, and Shepard Fairey.

 

Reference

Denial of responsibility! TechCodex is an automatic aggregator of Global media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, and all materials to their authors. For any complaint, please reach us at – [email protected]. We will take necessary action within 24 hours.
DMCA compliant image

Leave a Comment