Judge denies AI company’s move to toss music copyright claims

SAN JOSE, Calif. (CN) — A judge denied an artificial intelligence company’s push to dismiss a case built on claims of copyright infringement by a consortium of music publishing companies on Monday, upholding the plaintiffs’ assertion that the company’s use of lyrics through its AI was partially due to its removal of “copyright management information.”

U.S. District Judge Eumi Lee denied the motion to dismiss without hearing oral argument after Concord Music Group Inc. and other music publishers filed an amended complaint against San Francisco-based Anthropic, a tech company that produces a series of AI large language models called “Claude,” which provides answers in response to user prompts based on collected data.

Anthropic argued it didn’t know third-party users were committing infringement and didn’t directly financially benefit from Claude’s use of the group’s exclusive lyrics, among other claims.

Lee believed the plaintiffs’ arguments that Anthropic benefits financially because “Anthropic is paid every time … end users submit a request for publishers’ song lyrics, and it is paid again every time its Claude API generates output copying and relying on those lyrics,” she said in the denial.

Additionally, while training Claude, the music publishers claim Anthropic “collected Claude user prompts and output data, including specific infringing output copying lyrics,” said Lee. “And Anthropic supposedly became aware of the user prompts seeking such lyrics because, each time this occurred, the guardrails detected the users’ actions and Anthropic received a corresponding notice.”

Yet, according to the plaintiffs’ arguments, Anthropic knowingly didn’t stop Claude from using lyrics, and “at least some of these cofounders were aware that, during training, large language models tend to memorize ‘A LOT’ of data and then ‘regurgitate’ it,” said Lee.

“Taking these allegations together (as true), the court can plausibly infer that Anthropic had a reasonable basis to know that the dataset curation and training processes it engaged in, whereby it removed CMI (i.e., its intentional removal of CMI), concealed its own infringement,” she said.

Attorneys for the music publishers and Anthropic did not immediately respond to requests for comment.

This case isn’t the first time Anthropic has been on the legal hook for its AI. Earlier this year, Reddit sued the company over supposedly training Claude on its users’ data, and the company settled in September for $1.5 billion in a landmark case brought by book authors whose copyrighted works were downloaded and stored by the AI system. It was the largest settlement in the history of U.S. copyright cases.

On Wednesday, Lee will consider the plaintiffs’ request to file a second amended complaint.