AI company Anthropic has agreed to a $1.5 billion legal settlement with authors and publishers over allegations it illegally downloaded and stored millions of copyrighted books to train its AI models.
The agreement, which is the largest payout in U.S. copyright history, covers approximately 500,000 works and designates a payment of $3,000 per work. This settlement is part of a larger legal trend, with more than 40 lawsuits currently active nationwide between AI developers and copyright holders.
Some experts suggest the deal could establish a benchmark for future settlements, court decisions, or licensing fee structures. Music publishers are also monitoring the outcome with the hope of securing similar agreements for their industry. However, concerns have been raised that such high settlement costs could create a barrier to entry for smaller AI startups, favoring large, well-funded companies that can afford these payments.
For Anthropic, the $1.5 billion figure is relatively small compared to its recent $13 billion fundraising round and its estimated valuation of $183 billion. Previously, a judge had ruled that Anthropic’s use of copyrighted books for training its AI was “exceedingly transformative” and qualified as fair use.
The contested legal point remained how the company acquired the training materials, amid allegations of using pirated copies. A judge has postponed the approval of the settlement to conduct a further review, indicating the current agreement may not be finalized in its present form.
