Anthropic's $1.5 Billion Pirated Books Settlement Approved Amid New Patent Suit
A US federal judge has approved a landmark settlement between Anthropic and authors and publishers over the company’s use of pirated books to train its Claude AI chatbot. The agreement, worth $1.5 billion, marks a significant resolution in the growing wave of AI copyright litigation.
The settlement covers more than 482,000 books and will provide payments of approximately $3,000 per book to eligible authors and publishers. This means that rights holders can expect substantial compensation for their works being used without permission.
District Judge Araceli Martínez-Olguín ruled that the settlement provides ‘meaningful relief’ to those affected, with about 91% of covered works already claimed by authors and publishers. The judge’s decision is seen as a major victory for rights holders in the AI era.
The case was brought in 2024 by a group of authors led by bestselling novelist Andrea Bartz, who argued that Anthropic had used their works without permission in developing its AI systems. This move followed a significant court ruling that found training AI models on copyrighted books could qualify as fair use under US copyright law.
However, the same court also found that Anthropic had improperly obtained millions of books through pirate websites, creating the basis for legal claims against the company. The lawsuit highlights the complexities and challenges of navigating intellectual property rights in the age of large language models.
The agreement is regarded as the first major settlement among numerous copyright lawsuits currently facing AI developers over the use of published works in training their systems. This development comes amid growing concerns about the handling of copyrighted material by companies like Anthropic, which has sparked a wave of litigation and scrutiny from courts and regulators.
Despite this significant resolution, Anthropic is now facing another legal challenge. The University of Tennessee Research Foundation has filed a lawsuit against the company in federal court in Delaware, alleging that its AI systems infringe patents related to neural networks, machine learning, and neuroscience-inspired computing.
The foundation claims two patents developed by university researchers cover technologies used in Anthropic’s AI models. It is seeking unspecified financial damages and a court order barring further infringement. This new lawsuit adds to the broader legal scrutiny facing artificial intelligence companies as courts continue to define the boundaries of intellectual property rights in the AI era.
Anthropic has welcomed the outcome, highlighting the court’s earlier finding that AI training on books itself constituted fair use. The company expressed its satisfaction with the settlement and noted that the vast majority of eligible authors and publishers had claimed their share of the funds. This move is seen as a crucial step towards resolving the complex issues surrounding intellectual property rights in the age of large language models.
The patent suit filed by the University of Tennessee Research Foundation marks a significant escalation in the legal battles facing Anthropic. The case highlights concerns about the handling of patented technologies, which are increasingly being used in AI systems. This development underscores the need for companies like Anthropic to carefully navigate intellectual property rights and avoid infringing on patents.
The lawsuit is believed to be one of the first of its kind brought against Anthropic, adding to the growing list of legal challenges facing artificial intelligence companies. As courts continue to define the boundaries of intellectual property rights in the AI era, companies must ensure that they are respecting the rights of creators and innovators while developing their technologies.
The settlement approved by Judge Martínez-Olguín marks a significant milestone in the ongoing debate about AI copyright litigation. It highlights the need for companies like Anthropic to prioritize transparency and respect for intellectual property rights when using copyrighted material in training their systems.