Daniel Mercer
In a situation reminiscent of the aftermath of ChatGPT’s launch, music publishers are taking a bold stance with a new lawsuit against Anthropic. They claim it represents “one of the largest and most blatant ongoing thefts of intellectual property in history,” according to a report by Axios.
Initially, this legal action appears to be a curious move, primarily focused on song lyrics, and seems designed to provoke a lengthy legal battle—or perhaps a substantial settlement for the plaintiffs.
The lawsuit was filed late Friday in a federal court in Northern California. It targets Anthropic as a company, along with its co-founders, Dario Amodei and Benjamin Mann. The plaintiffs are seeking $150,000 for each alleged infringement, which could lead to damages totaling billions of dollars.
You may wonder, “Aren’t numerous music publishers already pursuing legal action against Anthropic?” The answer is yes; there is a widespread effort involving both major and independent publishers. Universal Music Group, Concord, and ABKCO initiated lawsuits in 2023, followed by BMG in March, and Round Hill Music earlier this month.
This latest lawsuit seems to adopt a dual strategy. As explained in the music industry publication Music Business Worldwide, part of the case relies on documents disclosed in a previous lawsuit: Bartz v. Anthropic. You might recall that this case involved a book author who reached a settlement with Anthropic last year, in which the company agreed to pay $1.5 billion. The allegations claimed that Anthropic had unlawfully downloaded thousands of books by accessing the Library Genesis (LibGen) and Pirate Library Mirror (PiLiMi) archives. Anthropic consented to destroy the pirated copies used or intended for training data.
While I am not a legal expert, it seems that unless lyric books were included in those alleged pirated collections, this part of the lawsuit mainly aims to establish a pattern of behavior that strengthens the plaintiffs’ argument. Torrenting e-books is an unusual and indirect method of acquiring song lyrics.
The lawsuit further contends that Anthropic infringed on the plaintiffs’ copyrights by scraping data from authorized lyric databases like MusixMatch and LyricFind. Allegedly, the AI model Claude can produce lyrics verbatim. The lawsuit provides examples, including:
“[…]such beloved songs as ‘Ain’t No Mountain High Enough,’ ‘All I Want for Christmas is You,’ ‘Eye of the Tiger,’ ‘Here Comes Santa Claus,’ and ‘Paper Rings.’”
This raises familiar, albeit legally ambiguous, issues similar to the notable New York Times lawsuit against OpenAI, Microsoft, and Perplexity. This case sidesteps any arguments regarding fair use related to training data and instead asserts that ChatGPT can allegedly reproduce verbatim and significant portions of paywalled New York Times articles. The legal rationale suggests that this could harm the New York Times, a key source of information, by allowing users to access that same content through a chatbot. The outcome of that case remains pending.
Gizmodo reached out to Anthropic for a comment but has not yet received a response.
Ethan Carter
