Leaked File Shows the Thousands of Artists Midjourney Has Allegedly Used to Train Its AI Models

It’s no surprise to hear that AI image generators rely on the work of real human beings to churn out content at blink-fast rate. But exactly which artists? According to reports, a Google Sheet revealed via X (formerly Twitter) and Bluesky contained a list of thousands of artists, time periods, styles, genres and techniques that Midjourney has allegedly been mining for its services.

Names include Yayoi Kusama, Damien Hirst, Andy Warhol, Francis Bacon, Banksy, HR Giger, and countless others. Even outside the traditional realms of art, card and video games like Magic: The Gathering and Riot Games have also noted of a breach in IP.

The list stems from a complaint originally filed a year ago against Midjourney, Stability AI and DeviantArt, which accounted 16,000 artists that the AI platforms have allegedly copied from, including the work of a six-year-old Hyan Tran, who created a drawing for the Seattle Children’s Hospital. The news comes off the heels of a landmark decision by the US Copyright Review Board, which ruled in September 2023, that images generated via AI platforms could not be protected by copyright law, due to the means of production.

Midjourney developers caught discussing laundering, and creating a database of Artists (who have been dehumanized to styles) to train Midjourney off of. This has been submitted into evidence for the lawsuit. Prompt engineers, your “skills” are not yourshttps://t.co/wAhsNjt5Kz pic.twitter.com/EBvySMQC0P

— Jon Lam #CreateDontScrape (@JonLamArt) December 31, 2023

While the leaked file was eventually scrubbed off social media, it can still be viewed in its entirety here.