Photo Credit: Igor Omilaev
A little over a month out from the one-year anniversary of its start, the copyright suit levied by major music publishers against Anthropic is heating up amid the AI giant’s renewed push for dismissal.
That push, music publisher plaintiffs including Concord and UMPG emphasized in their latest filing, actually marks the second attempt by Anthropic to dismiss the high-stakes case. As many know by now, the courtroom confrontation centers on alleged infringement stemming from the training process behind Anthropic’s Claude product.
The filing companies have pointed to the alleged presence of lyrics in the chatbot’s outputs – and claimed, among other things, that the alleged “massive copyright infringement” helps Anthropic to generate revenue and attract users.
Multiple twists and one time-consuming venue change later, Anthropic last month (again) fired back against the publishers’ push for a preliminary injunction blocking the continued use of lyrics in outputs and in future training.
And now, the music publishers themselves are taking the opportunity to refute Anthropic’s latest dismissal arguments as well as the appropriate motion.
Predictably, given the ultra-important case’s plodding nature, this 33-page refutation doesn’t break too much new ground. Instead, the publishers drove home that Anthropic’s dismissal motion is untimely in part because it arrived before a formal answer to the suit.
Running with the latter idea, the plaintiffs indicated that Anthropic had “deliberately contravened the Federal Rules” by ignoring purported warnings about the timing of its answer (or the lack thereof).
In short, the Amazon-backed AI mainstay is working “to gain a litigation advantage by prioritizing resolution” of the dismissal motion without first answering the complaint.
“When Anthropic answers the Complaint,” the publishers spelled out, “it will have to admit facts that it so far refuses to acknowledge directly, including that Anthropic copied Publishers’ lyrics when training Claude and made no effort to remove those lyrics from its training dataset despite its ability to do so.”
And from there, the publishers took aim at Anthropic’s specific dismissal arguments, which are looking to toss each claim save that involving direct infringement.
“Publishers are not required to name and date every instance of direct infringement to state a claim for secondary infringement,” the plaintiffs reiterated in part. “Publishers plausibly allege that Anthropic’s AI models respond to queries from users seeking copyrighted lyrics—including queries from Publishers’ investigators—by delivering those lyrics as requested.
“That allegation, without naming specific infringing users, is sufficient to set forth a valid claim for secondary copyright liability,” they proceeded.
Furthermore, dismissal would be “especially unwarranted” because the plaintiffs have yet to “discover from Anthropic what other third parties have requested from the Claude chatbot or APIs,” per the precedent-heavy legal text.
