A trio of major music publishers recently settled a lawsuit with Anthropic, an AI company backed by Amazon, regarding the use of lyrics to train their AI system. The agreement, approved by U.S. District Judge Eumi Lee, requires Anthropic to maintain existing safeguards that prevent their AI chatbot, Claude, from providing lyrics to songs owned by the publishers or generating new lyrics based on copyrighted material.
Anthropic has stated that Claude is not intended for copyright infringement and has processes in place to prevent such actions. They believe that using potentially copyrighted material in training AI models falls under fair use. The publishers, including Universal Music Group, Concord Music Group, and ABKCO, filed the lawsuit in Tennessee federal court, accusing Anthropic of infringing on copyrights by using lyrics from popular songs without permission.
The lawsuit highlights concerns about Anthropic undermining an existing market by using lyrics without consent or compensation. This legal action marks the first time a music publisher has taken legal action against an AI firm for incorporating lyrics into a large language model. The agreement reached mandates Anthropic to continue implementing guardrails in training their AI systems and allows publishers to raise concerns if these measures are not effective.
Anthropic reassures that their existing guardrails make it unlikely for Claude to produce substantial portions of copyrighted works. The court is expected to make a decision in the following months regarding a potential preliminary injunction that could prevent Anthropic from using lyrics owned by the publishers in future AI training.
Source link
