Anthropic Wins Major Copyright Settlement: $1.5 Billion to Authors and Publishers
In a landmark decision, Anthropic, the AI research lab, is set to disburse $1.5 billion following a class action copyright lawsuit. This significant settlement comes after a federal judge granted final approval, allowing the company to compensate authors and publishers who accused it of copyright infringement.
Key Updates on the Settlement
The final nod to the settlement was given by Judge Araceli Martinez-Olguin at the U.S. District Court for the Northern District of California on Monday. This follows Judge William Alsup’s preliminary approval last year, wherein he ruled that Anthropic had illegally downloaded and stored millions of copyrighted books.
Details of the Settlement
Under the terms of the settlement, authors and publishers are set to receive $3,000 for each of the approximately 500,000 works involved. While this is touted as the largest settlement of its kind in U.S. copyright history, it doesn’t necessarily bring cheer to all authors and creators.
Legal Precedents and Controversy
The controversy stems from how the legal situation was resolved. Judge Alsup supported Anthropic on the primary issue, stating that training AI models on copyrighted texts falls under fair use. This ruling marks a pivotal moment for the AI sector. However, it did not absolve Anthropic from how it acquired the copyrighted materials initially.
Anthropic sourced its training library from two primary avenues: legitimate purchases and scans of books, and downloads from illegal sites such as Library Genesis and Pirate Library Mirror. While the first method was deemed acceptable, the latter was ruled illegal, leading Judge Alsup to suggest the piracy issue could still be pursued in court. To sidestep a potential trial, Anthropic opted for a settlement.
Impact on Future Copyright Cases
Although this settlement closes the case against Anthropic, it leaves unresolved questions about copyright laws in the AI industry at large. Alsup’s ruling serves as a single district court decision. Anthropic’s choice to settle means the case will not proceed to higher courts as binding precedent.
Consequently, various judges nationwide remain free to interpret copyright laws in their own jurisdictions. This situation is exemplified by ongoing lawsuits against tech giants, including Google, Meta, Midjourney, and OpenAI. In fact, just last week, a group of publishers and authors—such as Hachette, Cengage, and Elsevier—filed a class action lawsuit against Google, alleging it also used their copyrighted works for training its AI platform, Gemini.
Conclusion: What Lies Ahead?
This landmark settlement may signify a cautious path forward for the AI sector regarding copyright issues, yet many creators remain unsettled. With continued legal challenges looming, the AI industry must navigate its complexities, balancing innovation with the rights of authors and publishers.
