Judge Lee denies UMG’s request to block AI company Anthropic from using song lyrics

Key Takeaways:

• Judge Eumi K. Lee denied UMG’s request for a preliminary injunction against Anthropic, ruling that publishers failed to demonstrate “irreparable harm”
• The court granted publishers favorable discovery terms, allowing them to gather more evidence from Anthropic’s Claude AI system
• The ruling doesn’t affect previous “guardrails” that prevent Claude from reproducing copyrighted lyrics in its outputs to users

A federal judge in California has rejected Universal Music Group’s request to block AI company Anthropic from using song lyrics to train its AI models. In a ruling issued on Tuesday, March 25, Judge Eumi K. Lee of the U.S. District Court for the Northern District of California denied the publishers’ motion for a preliminary injunction, marking a significant development in the ongoing copyright battle between music rights holders and AI developers.

The lawsuit, filed in 2023 by UMG, Concord, and ABKCO, alleges that Anthropic trained its AI chatbot Claude on lyrics from at least 500 songs by artists including Beyoncé, the Rolling Stones, and the Beach Boys without permission.

However, Judge Eumi K. Lee determined that the publishers failed to demonstrate “irreparable harm” – a necessary condition for granting such an injunction.

The court noted that the existence of licensing agreements between other AI developers and content owners suggests that any potential harm could be compensated through monetary damages rather than requiring immediate injunctive relief.

Judge Lee also expressed concerns about the broad scope of the requested injunction, which would have covered not just the 500 works identified in the lawsuit but potentially “hundreds of thousands” of songs.

Despite this setback for the music publishers, they did secure a partial victory. The court issued two discovery orders allowing UMG and its fellow publishers to gather more evidence from Anthropic, including a “statistically significant sample” of Claude prompt and output records from September 2023 to March 2024.

Importantly, this ruling doesn’t affect a previous agreement approved by the court in January 2025, where Anthropic implemented “guardrails” to prevent Claude from reproducing copyrighted lyrics in its responses to users. The case continues as part of growing legal battles over AI and copyright, with potential far-reaching implications for both creative industries and AI technology development.