Source: JD Supra

JD Supra reports that Judge Eumi K. Lee of the U.S. District Court for the Northern District of California denied Anthropic PBC’s second motion to dismiss in Concord Music Group Inc. v. Anthropic PBC. The decision allows a group of major music publishers to proceed with claims of copyright infringement and violations of the Digital Millennium Copyright Act (DMCA). The publishers allege that Anthropic’s Claude AI model reproduces copyrighted song lyrics without authorization and intentionally removes copyright management information (CMI) from its training data. The court found the allegations plausible, noting that Claude was trained on vast datasets known to contain protected content.

A notable aspect of the ruling is the court’s mention of Anthropic’s own "guardrails" as potential evidence that the company was aware of infringing outputs but proceeded anyway. Additionally, the judge accepted the threshold for a DMCA claim regarding the alleged use of a "Newspaper" algorithm designed to strip author names and notices. This case highlights a growing judicial trend of allowing infringement-based challenges against generative AI to bypass initial dismissal stages. As the litigation moves into discovery, it will provide critical insights into whether AI developers can be held liable for the internal processing of intellectual property during the training of large language models.