Artists Battle AI Companies Over Use of Creative Works
Authors, musicians, and illustrators are taking tech giants to court over AI training data they never agreed to provide.
Key points
- Kirk Wallace Johnson found his books used to train AI without his consent.
- Illustrator Sarah Andersen filed a class action against Stability AI and Midjourney in January 2023.
- Google faces a lawsuit over YouTube data used to train its Lyria music AI.
- Andrea Bartz is lead plaintiff in a copyright suit against Anthropic.
- Artists say AI companies fundamentally disrespect creative work.
When The Atlantic published a searchable dataset of books used to train AI, author Kirk Wallace Johnson looked up his own name and found it. His nonfiction titles, including 'The Feather Thief', had been scraped and fed to a model without his knowledge. He describes feeling "anger over the brazenness of the theft, worry over what this means for writers, and a healthy thirst for revenge" on the corporations that profited. He contacted Susman Godfrey, the law firm handling the case against Anthropic, because he saw the suit as a "middle finger on behalf of everyone that has tried to create something."
Why are artists suing AI companies?
Creators say their work was taken without permission and used commercially. Johnson is one of dozens of authors, musicians and illustrators who've filed suit, mostly on copyright grounds, though some have pursued terms of service violations instead. The cases have run for years in some instances and settled quickly in others.
What are the key cases?
Illustrator Sarah Andersen, joined by Karla Ortiz, Kelly McKernan and others, filed a class action against Stability AI, Midjourney, DeviantArt and Runway AI in January 2023 and the case is still moving through the courts. As we reported on 20 July 2026, Sony's fresh copyright filing against Udio shows the music industry is pressing on a parallel front. Musician Sam Kogon leads a separate suit accusing Google of using its Content ID system and YouTube data to train Lyria and ProducerAI while breaching its own terms of service. Google has filed a motion to dismiss, arguing its terms grant it broad rights over uploaded content.
| Case | Plaintiff | Defendant | Date Filed | Allegation |
|---|---|---|---|---|
| Andersen v. Stability AI | Sarah Andersen | Stability, Midjourney | Jan 2023 | Copyright infringement |
| Bartz v. Anthropic | Andrea Bartz | Anthropic | - | Copyright violation |
| Kogon v. Google | Sam Kogon | - | TOS violation |
What does this mean for the future of AI and art?
The verdicts will set the practical boundaries of what training data companies can use. Independent creators carry the most risk: novelist Richard Kadrey's earlier filing against Meta put it plainly, arguing that AI-generated books might not threaten established names but could stop the next generation of writers from finding an audience at all.
Common questions
What is fair use in this context?
Fair use allows limited, unpaid use of copyrighted material for criticism or education, but artists argue that commercial AI training doesn't meet that bar.
How could this affect independent artists?
Independent artists face the sharpest threat. If AI output floods the market, the mid-list writers and working musicians who depend on steady income have the least cushion.
What rights do companies have over uploaded content?
Terms of service often grant platforms broad licences, but how far those licences extend to AI training is precisely what courts are being asked to decide.
