Details about a secretive effort by AI company Anthropic to scan millions of books for its artificial intelligence models have come to light through unsealed court documents. The project, known internally as Project Panama, aimed to destructively scan books to feed data into AI tools like its popular chatbot, Claude.

What the Documents Reveal

According to internal planning documents, Anthropic spent tens of millions of dollars acquiring books and slicing off their spines to scan the pages. The company reportedly wanted to keep the project under wraps, stating, “We don’t want it to be known that we are working on this.”

The revelations emerged from more than 4,000 pages of legal filings in a copyright lawsuit brought by book authors. Anthropic settled the case in August for $1.5 billion, but the unsealed documents shed light on the company’s aggressive tactics to obtain books without authors’ consent.

Broader Implications

Anthropic isn’t the only company facing allegations of using copyrighted material without permission. Meta, Google, Microsoft, and OpenAI, the maker of ChatGPT, are also dealing with similar lawsuits from authors, artists, and news outlets.

Court filings allege that Anthropic, Meta, and others acquired books in bulk, sometimes by downloading pirated copies. Internal messages from Meta employees reportedly raised concerns about the legality of the practice. In one instance, a December 2023 email mentioned the approval of downloading books after “escalation to MZ,” believed to refer to Meta CEO Mark Zuckerberg. Meta has declined to comment on the allegations.

Authors Fight Back

Authors have been vocal about their frustration, arguing that AI companies should seek permission before using their work. The lawsuits highlight a growing tension between creators and tech companies as artificial intelligence continues to advance.

Here in Oklahoma, where folks value hard work and creativity, the idea of someone’s work being used without permission doesn’t sit well with many. Local writers and artists are watching these cases closely, wondering what it might mean for their own livelihoods.

What’s Next?

While Anthropic and other companies face mounting legal challenges, the unsealed documents have sparked broader conversations about copyright in the digital age. The outcomes of these lawsuits could set important precedents for how AI companies use creative works in the future.

For now, authors and creators are urging companies to be more transparent and to respect intellectual property rights. As these cases unfold, it’s clear that the intersection of technology and creativity is entering uncharted territory.