A US judge has officially approved a $1.5 billion settlement in a copyright lawsuit filed against AI company Anthropic. This decision brings to a close one of the most significant copyright class action cases in the United States, stemming from allegations that Anthropic used copyrighted books without permission to train its Claude chatbot. The settlement, initially agreed upon in September 2025, was formally approved this week.
The legal battle began in 2024 when authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson initiated a lawsuit against Anthropic. They contended that the company had incorporated pirated books into the training data for its AI model. While a judge ruled in June 2025 that using authors' work for AI training could fall under US copyright law's 'fair use' doctrine, Anthropic faced continued legal scrutiny because many of the books used were reportedly acquired online without proper authorization.
The approved settlement covers approximately 500,000 distinct works, with eligible authors and publishers set to receive $3,000 per pirated work. According to Anthropic, 91% of those entitled to compensation have already filed claims.
Some of the affected authors expressed concerns that the $1.5 billion payout was insufficient, particularly given that Anthropic's 'found' library initially exposed the company to potential statutory damages of up to $150,000 per work. This perspective is further emphasized by Anthropic's reported valuation of $965 billion in May and its prediction of reaching its first profitable quarter this year. For these reasons, the settlement amount has been perceived by some as a relatively minor consequence for the AI company.
Despite the concerns about the settlement's scale, the resolution offers a measure of recompense for creators in the rapidly evolving landscape of AI development. Similar lawsuits targeting the AI training practices of major companies such as Google, OpenAI, and Meta are ongoing. The outcome of these future cases may set further precedents for copyright protection and fair compensation in the age of generative AI.