Skip to content

AI industry alarmed by largest certified copyright infringement class action lawsuit

The AI industry is facing an unprecedented legal challenge after a federal court in California certified the largest copyright infringement class action lawsuit to date. Three authors accuse the AI company Anthropic of using copyrighted works without permission to train its AI model Claude. Industry associations are now calling on an appeals court to halt the certification of the lawsuit, as it could pose a financial threat to the entire AI industry.

Anthropic has appealed the decision, criticizing that the presiding judge, William Alsup, did not conduct a thorough analysis of the potential plaintiff group. The lawsuit potentially includes up to seven million copyright holders whose works were published over decades. If the appeal is rejected, Anthropic faces damages claims of hundreds of billions of dollars, as copyright infringements in the US can be punished with up to $150,000 per work.

The lawsuit, filed in 2024, is based on allegations that Anthropic downloaded millions of books from piracy libraries such as LibGen and PiLiMi to train Claude. While Alsup recognized the use of legally acquired books for AI training as "fair use," he ruled that downloading pirated copies is not covered by this exception. A trial to clarify the amount of damages is scheduled for December 2025.

The decision has far-reaching implications for the AI industry. Industry representatives warn that a plaintiffs' victory could fundamentally change the business models of AI companies and hinder innovation, as the procurement of legal training data would become more expensive and complex. Anthropic argues that the certification of the class action lawsuit was rushed and did not sufficiently consider the complexity of copyright issues. The appeals court is now tasked with clarifying whether the "fair use" doctrine should be evaluated holistically or on a process-by-process basis for AI training, and whether the illegal acquisition of data excludes the "fair use" defense.

The lawsuit is supported by the Authors Guild, which sees the decision as an important step for protecting authors' rights. At the same time, AI companies emphasize that the use of copyrighted works for transformative purposes such as AI training is covered by copyright. The outcome of the appeal process could set groundbreaking precedents for the entire industry.

Source: Ars Technica

author avatar
LabNews Media LLC
The Editors in Chief of labnews.ai are Marita Vollborn and Vlad Georgescu. They are bestselling authors, science writers and science journalists since 1994.More details about their writing on X-Press Journalistenbüro (https://xpress-journalisten.com).More Info on Wikipedia:About Marita: https://de.wikipedia.org/wiki/Marita_Vollborn About Vlad: https://de.wikipedia.org/wiki/Vlad_Georgescu
LabNews Media LLC

LabNews Media LLC

The Editors in Chief of labnews.ai are Marita Vollborn and Vlad Georgescu. They have been bestselling authors, science writers, and science journalists since 1994.More details about their writing at X-Press Journalistenbüro (https://xpress-journalisten.com).More Info on Wikipedia:About Marita: https://de.wikipedia.org/wiki/Marita_Vollborn About Vlad: https://de.wikipedia.org/wiki/Vlad_Georgescu