Sony Music Publishing and Warner Chappell Music have filed a lawsuit against Anthropic in the U.S. District Court for the Northern District of California, seeking damages for “tens of thousands” of copyrighted works. The two publishers called it “one of the largest and most egregious ongoing thefts of intellectual property in history.”
The damages sought reach up to $150,000 per work, plus $25,000 for each instance where copyright management information was removed. If the court awards the maximum statutory penalties, total damages could reach several billion dollars.
In addition to suing Anthropic as a corporate entity, the lawsuit also names co-founders Dario Amodei and Benjamin Mann individually. Specifically:
- Mann is accused of using BitTorrent to download more than 5 million pirated books.
- Anthropic employees allegedly downloaded at least 2 million more books from Pirate Library Mirror.
- Anthropic is accused of scraping song lyrics from Musixmatch and LyricFind—platforms that pay licensing fees to publishers.
Songs identified in Anthropic’s training data include “Ain’t No Mountain High Enough” (Marvin Gaye), “Livin’ on a Prayer” (Bon Jovi), “September” (Earth, Wind & Fire), “Hallelujah” (Leonard Cohen), and “Paper Rings” (Taylor Swift).
This is the latest in a series of copyright lawsuits targeting Anthropic. The company recently settled a lawsuit with the publishing industry for $1.5 billion, while also facing lawsuits from Universal Music Group, Concord, ABKCO, BMG, and Round Hill Music.
This development raises a major question for the entire AI industry: as major labs face continuous lawsuits over training data, is the “scrape first, ask permission later” model still sustainable? If these music lawsuits prevail with maximum penalties, it will serve as a wake-up call—not only for Anthropic, but for every AI company relying on copyrighted data to train models.