Overview
- Judge William Alsup reaffirmed in a July 16 opinion that training Anthropic’s Claude model on copyrighted books qualifies as fair use under 17 U.S.C. § 107
- Alsup held that Anthropic’s central repository of roughly seven million books downloaded from LibGen and PiLiMi does not enjoy fair use protection
- The court noted that illicit acquisition and long-term storage of pirated works could expose Anthropic to billions of dollars in potential damages
- Certification empowers Andrea Bartz, Charles Graeber and Kirk Wallace Johnson to represent all U.S. authors whose works were allegedly pirated for AI training
- Legal experts say the ruling may influence parallel AI copyright cases, including multidistrict litigation in the Southern District of New York