Anthropic Reaches Settlement in AI Copyright Lawsuit with Authors

Anthropic, an artificial intelligence company, has reached a preliminary settlement in a class action lawsuit filed by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson. The authors alleged that Anthropic used their copyrighted works without permission to train its AI models. The settlement is expected to be finalized by September 3, with further details forthcoming.

In June, U.S. District Judge William Alsup ruled that Anthropic's use of the authors' books for AI training constituted "fair use" under U.S. copyright law. However, the judge found that Anthropic's acquisition of some works from online "shadow libraries" like LibGen amounted to piracy, allowing the authors to pursue a class action trial on this issue. The trial was scheduled to begin in December.

Statutory damages for such piracy start at $750 per infringed work. Given that Anthropic's library reportedly contained approximately 7 million works, the company faced potential penalties amounting to billions of dollars. Legal experts noted the significant risk for Anthropic, with potential damages reaching "doomsday" amounts.

The settlement terms have not been disclosed, and it remains to be seen how the author class will respond once details are unveiled. This case is among several copyright-related legal challenges facing AI companies, highlighting the ongoing debate over the use of copyrighted material in AI training.