[AI Summary]: Dan Cohen analyzes the landmark $1.5+ billion settlement in Bartz v. Anthropic, where AI company Anthropic agreed to compensate authors of approximately 500,000 books used for training Claude. While Judge Alsup ruled that training AI on purchased books is fair use, he found potential copyright violations in downloading books from unauthorized sources. Cohen argues that despite the financial victory, the settlement may paradoxically strengthen major AI companies by creating a billion-dollar entry barrier for competitors while failing to address authors’ deeper desires for recognition, proper citation, and readers. He suggests that rather than cash payments, authors need AI systems that provide proper attribution, encourage full reading of texts, and serve as discovery tools rather than substitutes for human-authored works.
- Author: Dan Cohen
- Date: September 5, 2025
- Source: Humane Ingenuity newsletter