Music publishers suffered a setback in their copyright case against Anthropic on November 4, 2026, when Judge Eumi K. Lee of the U.S. District Court for the Northern District of California dismissed the direct copyright infringement claim pending against the company's co-founder and CEO, Dario Amodei. As first reported by Courthouse News and MLex, the dismissal came with leave to amend — meaning publishers can rewrite the claim and try again — and left the core of their case against Anthropic itself untouched.
The ruling, documented in the court's November 4 minute entry and analyzed in detail by Unite.AI, is narrow by design. It addresses one theory of personal liability against one defendant in Concord Music Group, Inc. v. Anthropic PBC, a case that also names Anthropic co-founder Benjamin Mann alongside the company. For more context on this story, see our ongoing latest AI developments.
Why Amodei Was Dismissed — For Now
Amodei's August 3, 2026 motion targeted only the allegation that he directly infringed copyrights by torrenting the publishers' works. His argument was precise: the complaint attributes the actual downloads to Mann and other Anthropic employees, while the allegations that Amodei directed, approved, knew of or encouraged that conduct describe something different — contributory liability, not personal copying. Direct infringement requires unlawful copying by the defendant himself; contributory infringement addresses knowingly inducing or materially contributing to someone else's infringement.
Judge Lee agreed to dismiss the direct claim, though without prejudice: the publishers have leave to amend, giving them a chance to plead the direct-copying allegation against Amodei more specifically. His motion had sought dismissal with prejudice, which would have ended that claim entirely. The contributory theory against him was never challenged and therefore remains in the case.
The Claims Still Standing
The ruling leaves the architecture of the publishers' lawsuit largely intact. Their direct-infringement claim based on torrenting still stands against Anthropic — Amodei's motion did not address it — as does the separate claim built on Claude's training and outputs. Anthropic's own August 3 motion sought to kill two other pieces: the portion of the direct-infringement claim based on Claude's outputs, and the entire Digital Millennium Copyright Act claim concerning copyright management information. That partial motion remains unresolved.
The CMI claim hinges on identifying information conveyed alongside a work — its title, songwriter and copyright owner. Anthropic argues the publishers' allegations concern lyrics rather than complete copyrighted musical compositions, and has asked for dismissal with prejudice on that count. The company also contends the complaint supplies no specific example of an infringing Claude output and does not distinguish which works were allegedly copied during training versus reproduced in responses.
A Dispute Over Pleading Standards, Not Proof
Unite.AI's analysis of the docket stresses what the ruling does not decide: whether any particular Claude response has actually been shown to copy a song. The live question is whether the publishers pleaded their output and CMI theories with enough specificity to survive dismissal at this stage.
The publishers' second amended complaint, filed May 21, 2026, identifies 20,517 compositions and alleges that Anthropic copied lyrics without permission for training, generated verbatim and near-verbatim copies in Claude outputs, and deployed guardrails that covered only a limited set of songs rather than the broader catalog at issue. Their theory links the stages: initial copying allegedly encoded the lyrics into the model, and further training and fine-tuning allegedly produced copies in response to user requests.
In their August 31, 2026 opposition, the publishers argued they need not identify specific infringing outputs at the pleading stage, and that records of third-party users' prompts and Claude's corresponding responses sit in Anthropic's possession and control. On CMI, they argued that scanning and torrenting songbooks and sheet music copied complete compositions, and that extraction algorithms stripped identifying information during training.
What Happens Next
Judge Lee ordered the parties to meet and confer on whether to resolve Anthropic's outstanding partial motion through additional briefing or by amending the complaint, and to file a joint status report by November 11, 2026. MLex characterized the hearing's aftermath as the judge doubting aspects of the music publishers' claims — framing that suggests the amendment path may not be easy.
The Bigger Picture for Anthropic
The decision lands in a year in which copyright has been expensive for Anthropic. In July 2026, a judge approved a $1.5 billion settlement over pirated books used to train Claude — one of the largest copyright resolutions in history, per AP News reporting at the time. Against that backdrop, the Concord case remains the sharper threat on a specific front: it targets not just training but what the model emits, an output-based theory few AI cases have pressed this far.
For now, though, the personal-liability angle against Anthropic's CEO has been pared back, and the company's core defenses remain untested. The next checkpoint arrives November 11, when the parties must tell the court how they intend to proceed.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
