Matt Larimer, July 30th, 2026

According to a series of reports from Futurism, the artificial intelligence company Anthropic settled a $1.5 billion lawsuit in August of 2025 which revealed the company was destroying millions of books to train its large language model Claude in a secret initiative it named Project Panama.

The Anthropic website describes their governance as, “a Public Benefit Corporation, whose purpose is the responsible development and maintenance of advanced AI for the long-term benefit of humanity.“

However, internal documents retrieved by the Washington Post showed that the company knew there would be ethical concerns about its action stating that, “We don’t want it to be known that we are working on this.”

Futurism reported that Anthropic would source books from bulk distributors, use hydraulic cutting machines to remove the spines of the books, allowing them to be processed by high speed scanners, then schedule recycling companies to pick up the destroyed remnants. 

The judge in the case ruled that the book destruction was technically legal because the company had purchased the books and was not distributing their contents for profit but processing them through a “transformative” process. 

The $1.5 billion settlement resulted from its previous practice of downloading millions of pirated books from digital libraries like LibGen and Pirate Library Mirror. Internal documents indicated that Anthropic knew these libraries were violating copyright laws.