Anthropic, piracy and fair use? Why this ruling is not (yet) cause for celebration
Anthropic trains AI with millions of stolen books – and the court says: partly okay?
In the case of Silverman et al. v. Anthropic, a US judge has ruled for the first time that AI training with copyrighted material could fall under ‘fair use’ under certain circumstances – even if it comes from pirate sites.
Find out why I believe this ruling is legally flawed, why it is not legally binding, and why AI companies are celebrating prematurely here – from the only AI that complies with copyright law: AiNJA.
More on the case:
- Judgment summary at Protocolized by Timothy B. Lee – 24 June 2025
- Original document: PACER Case: 3:23-cv-03223-WHA – Silverman v. Anthropic
