More than 100 authors have filed a lawsuit against Anthropic seeking over $75 million in damages, alleging the AI company pirated their copyrighted books to train its Claude AI model. The case, filed on June 17, 2026 in the US District Court for the Northern District of California, represents a new legal front that focuses not on whether AI training constitutes fair use, but on how the training data was obtained in the first place. The lawsuit, titled Shakespeare v. Anthropic, is led by British sociologist Thomas William Shakespeare, with the plaintiffs primarily being authors who opted out of a previous settlement in the Bartz v. Anthropic case. The complaint specifically alleges that Anthropic used BitTorrent to download works from Library Genesis and Pirate Library Mirror, stored books in a central library, and uploaded copies to other BitTorrent users during the training process.
The lawsuit targets some of the most prominent works in modern literature, including Oprah-endorsed New York Times bestseller "Get Good with Money" by Tiffany Aliche and the international bestseller "Like Water for Chocolate" by Mexican author Laura Esquivel. The plaintiffs include Atari co-founder Nolan Bushnell, Academy Award-nominated JFK screenwriter Zachary Sklar, and Newbery Medal winner Donna Barba Higuera, as well as other writers and rights holders listed in the filing. The authors are demanding $150,000 per pirated piece, arguing that copyright law allows statutory damages of up to $150,000 per willfully infringed work. As reported by The New York Post, the plaintiffs believe existing settlements undervalue their works, arguing that copyright law allows statutory damages of up to $150,000 per willfully infringed work. The practical implications extend beyond legal theory, as the case raises questions about whether dataset acquisition, retention, and redistribution records can survive discovery, making copyright risk around model training an evidence-management problem for AI teams.
The lawsuit includes prominent authors and rights holders among its plaintiffs, including Nolan Bushnell, Laura Esquivel, Tiffany Aliche, and Donna Barba Higuera, as well as other writers and rights holders listed in the filing. The case represents a significant shift in copyright litigation strategy, moving from model outputs to dataset acquisition, retention, and alleged redistribution issues for model builders. As reported by industry experts, the practical controls that model teams should implement include source-level license records, crawler and torrent exclusion policies, corpus manifests, retention rules, deletion logs, and review checkpoints before data moves into training or evaluation stores. The case reinforces that copyright compliance moves from policy documents into required dataset observability and data-governance tooling, with the operational question being whether copyright compliance becomes part of ML infrastructure rather than a legal afterthought.
The new lawsuit adds to mounting legal challenges facing Anthropic. The company already faces a separate class action filed in June over its Claude Max subscription plans, which targets the company on completely different grounds. In that earlier suit, plaintiff Karl Kahn alleged that advertised usage boosts collapsed under hidden caps, seeking refunds for subscribers since the plans launched in 2025. The trajectory from Bartz to Shakespeare also suggests that Anthropic's legal exposure is compounding rather than resolving, with settling one case for $1.5 billion only to face another for $75 million from authors who thought the first settlement was insufficient. The next signal will be whether defendants challenge the complaint on pleadings, settlement scope, or fair-use grounds, and whether the court treats the alleged BitTorrent distribution theory separately from training-use arguments.
The copyright case carries far heavier financial stakes than previous legal challenges. Anthropic previously settled a landmark class action for approximately $1.5 billion, paying authors around $3,000 each for an estimated 500,000 pirated books covered under that agreement. The earlier Bartz v. Anthropic case resulted in a proposed $1.5 billion settlement covering approximately 480,000 works, with preliminary approval granted in 2025. Authors who stayed in that settlement stood to receive roughly $3,000 to $3,100 per work after legal fees, and claims were filed for over 90% of eligible works. Some authors chose to opt out of that settlement, retaining the right to pursue individual claims for far larger per-work damages, with this new lawsuit representing their collective action. As part of that settlement, Anthropic said it will destroy downloaded copies of books the authors accused it of pirating, and under the deal it could still face infringement claims related to material produced by the company's AI models.
