What's Happening?
Anthropic has agreed to pay $1.5 billion to settle allegations of illegally downloading hundreds of thousands of books from pirate databases Library Genesis and Pirate Library Mirror to train its large language models. The settlement follows a ruling by U.S. District Judge William Alsup allowing a class action lawsuit by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson to proceed. The agreement includes destroying all pirated copies used in training. Attorneys for the authors describe the settlement as the largest publicly reported payout in U.S. copyright litigation history. The Authors Guild and the Association of American Publishers hope the settlement will encourage AI companies to adhere to copyright laws.
Why It's Important?
This settlement marks a significant precedent in the AI industry, emphasizing the need for companies to respect copyright laws when training AI models. It highlights the growing tension between technology companies and content creators over intellectual property rights. The agreement could lead to more stringent regulations and practices in the AI sector, potentially affecting how AI companies source data for model training. Authors and publishers stand to benefit from increased protection and compensation for their works, while AI companies may face higher costs and legal scrutiny.
What's Next?
A hearing is scheduled for September 8 to seek preliminary approval of the settlement, with final approval expected next year. The parties involved are working on a final list of works included in the agreement, due by October 10. Authors and publishers will be able to search for their works in a database, and Anthropic will add $3,000 for each additional title if the final number exceeds 500,000. Notices will be sent out this fall to instruct class members on filing claims, with the option to opt out and pursue individual lawsuits.
