Anthropic has filed a strong opposition to major music publishers' request for preliminary injunctions against its AI chatbot Claude, citing robust content safeguards and fair use arguments.

The publishers, including Concord and Universal Music, seek two main injunctions: removing protected works from Claude's training data and preventing the chatbot from outputting copyrighted lyrics.

Key points from Anthropic's opposition:

• Training data usage constitutes fair use under copyright law
• Monetary damages would adequately compensate publishers if they prevail
• Claude learns from "trillions of tiny textual data points"
• The company has implemented extensive safeguards against copyright infringement
• No evidence of ongoing market harm from alleged infringement

Anthropic argues that using copyrighted lyrics as part of a massive training dataset is "transformative" under fair use doctrine. The company's co-founder Jared Kaplan provided detailed testimony about Claude's training process in a supporting declaration.

The AI company maintains that its safeguards make future copyright violations unlikely, stating there's "no reasonable expectation that the complained-of activity is likely to recur."

Reports suggest a federal judge may dismiss a substantial portion of the publishers' lawsuit in the near future. The case continues under docket number 5:24-cv-03811.

Related Articles

Previous Articles