Notitia Platform 
Court Rules Anthropic Can Train AI Without Book Permissions, Faces Trial for Piracy
On June 24, 2025, U.S. District Judge William Alsup ruled that Anthropic’s use of books to train its Claude large language model (LLM) constitutes “fair use,” a win for the AI industry. However, the judge found that Anthropic’s storage of over 7 million pirated books in a “central library” infringed authors’ copyrights, ordering a December trial to determine damages.
Key Rulings
- Fair Use for AI Training: Alsup deemed Anthropic’s training of Claude with books by Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson as “transformative,” aligning with copyright law’s aim to foster creativity and scientific progress.
- Copyright Violation: Anthropic’s copying and storage of pirated books was ruled not fair use, violating authors’ rights. Statutory damages could reach $150,000 per work for willful infringement.
- Trial Scheduled: A December 2025 trial will assess damages owed by Anthropic, backed by Amazon and Google, for the piracy.
Background
The ruling stems from a 2024 class-action lawsuit by Bartz, Graeber, and Johnson, alleging Anthropic used pirated versions of their books without consent or compensation to train Claude. Anthropic’s spokesperson welcomed the fair use ruling, emphasizing its role in advancing AI innovation.
Broader Context
The decision reflects growing tensions between AI companies and content creators:
- BBC vs. Perplexity: On June 20, 2025, the BBC threatened legal action against AI search engine Perplexity for allegedly training its model with BBC content, demanding compensation and an end to scraping. The BBC cited 17% of Perplexity’s responses using its sources as inaccurate or lacking context.
- Industry Backlash: Publishers like The New York Times and Dow Jones have issued similar legal threats. Perplexity launched a revenue-sharing program to address concerns, but friction persists.
Implications
- AI Industry: The fair use ruling supports AI firms’ ability to train models with publicly available data, potentially accelerating innovation.
- Publishers: The piracy ruling and upcoming trial signal stronger protections for creators, with significant financial consequences for unauthorized use.
- Public Trust: Unauthorized scraping risks undermining trust in media outlets like the BBC, whose reputation for impartiality is critical.
Conclusion
Judge Alsup’s ruling balances AI innovation with copyright protections, allowing Anthropic to train Claude without book permissions but holding it accountable for piracy. The December trial could set a precedent for damages in AI-related copyright disputes. As tensions rise, with cases like BBC vs. Perplexity, the industry faces pressure to develop ethical data practices and revenue-sharing models to maintain trust and compliance.