BLOG
Class‑Action Lawsuit Against Anthropic
In a class‑action lawsuit against Anthropic, authors accuse the AI lab of training its Claude model on millions of pirated books downloaded from shadow libraries like LibGen and PiLiMi. While U.S. District Judge William Alsup recently held that training a model on lawfully acquired books may constitute “fair use” under section 107 of the U.S. Copyright Act, he distinguished that principle from the alleged unlawful acquisition and storage of copyrighted works. The plaintiffs argue that Anthropic’s bulk downloading and archiving of pirated material infringed authors’ exclusive rights to reproduce and distribute their works (17 U.S.C. §§ 106, 501) and could result in statutory damages well into the billions. A separate trial will determine liability and damages.
At AryaTech Heidarpour, we monitor developments in AI and copyright law and advise startups, investors and platforms on how to mitigate litigation risk. Whether you’re training models on proprietary data or navigating cross‑border compliance, our boutique firm can help you structure agreements, assess fair‑use defences and resolve disputes efficiently.
Source:
Anthropic cofounder and CEO Dario Amodei in May. The class action lawsuit against Anthropic centers on the company’s use of potentially pirated books to train its large language model.
