IT lexicon › Chat & Social Media › Bartz v. Anthropic

Bartz v. Anthropic

På svenska → Updated: 2026-11-02

The first major US ruling on training AI on books, which ended in a $1.5 billion settlement.

Three authors sued Anthropic in 2024 for training Claude on their books. In June 2025 Judge William Alsup split the case in two. Training a model on books the company had bought and scanned was fair use — "exceedingly transformative", in his words. Downloading millions of pirated copies from LibGen and similar sources and keeping them in a permanent library was not, whatever they were later used for.

The case was certified as a class action, and with statutory damages per work the exposure was enormous. In September 2025 the parties settled for $1.5 billion, roughly $3,000 per book — the largest copyright settlement in US history, and a clear price tag on skipping the purchase.

← Back to the lexicon