Overview

• U.S. District Judge William Alsup ruled that training Anthropic’s Claude model on copyrighted books without permission qualifies as fair use under U.S. law.
• Alsup determined that storing authors’ works in an internal “central library” exceeded fair use and infringed their copyrights.
• A trial has been scheduled to determine damages for the company’s retention of pirated book copies acquired for that library.
• Alsup described the AI’s training process as “exceedingly transformative,” likening it to a reader using source material to craft new works.
• This landmark ruling is the first to address fair use in generative AI and could influence similar lawsuits against other AI developers.