Anthropic, backed by Amazon and Google, has submitted a 40-page response opposing the preliminary injunction requested by music publishers Universal Music, Concord, and ABKCO in their ongoing copyright infringement lawsuit.

The publishers claim Anthropic infringed their copyrights by using their compositions without permission to train their AI systems and allege that Anthropic's Claude AI assistant copied song lyrics verbatim without attribution.

Key points from Anthropic's response:

• Training large language models (LLMs) on copyrighted content constitutes fair use
• Publishers haven't demonstrated irreparable harm warranting an injunction
• Anthropic attempted to implement additional safeguards and cooperate with publishers before the motion
• Requiring licenses for AI training would make general-purpose AI tools impossible due to the vast amount of data needed
• The company argues the venue is inappropriate as they have no relevant connection to Tennessee

Anthropic challenges the publishers' standing by noting that:

• Similar cases nationwide proceed without preliminary injunctions
• Publishers waited months before notifying Anthropic of alleged violations
• The alleged infringement may have been caused by the publishers' own prompts to Claude

The case highlights growing tensions between AI companies and content creators over training data usage and copyright implications. Anthropic maintains that money damages would sufficiently compensate publishers if they prevail, making an injunction unnecessary.

This legal battle represents a crucial test case for determining how copyright law applies to AI training data and could set important precedents for the AI industry's future.

Related Articles

Previous Articles