Overview

• Judge William Alsup held that using copyrighted works to train the Claude AI model is fair use because it transforms original texts into new content under Section 107.
• The court found that Anthropic’s storage of over seven million pirated book copies in its “central library” infringed authors’ rights and fell outside fair use.
• A December trial will now determine damages for the pirated materials, with statutory penalties potentially reaching $150,000 per work for willful infringement.
• Anthropic defended its approach by arguing that digitizing purchased books for AI training promotes innovation consistent with copyright law.
• Observers project that Alsup’s ruling will influence other AI copyright lawsuits, though appeals are anticipated.