A US federal judge sided with the artificial intelligence company Anthropic over its move to train models on published books without the authors’ permission.
It’s the first time that a court has backed Big Tech’s argument that the use of copyrighted material in training data constitutes “fair use,” like reviewing or parody.
The ruling is a blow to the authors, artists, and publishers who have brought lawsuits against AI companies: While there is no guarantee that other judges will follow suit, it sets a precedent.
Fair use is a “notoriously finicky” point of US copyright law, TechCrunch reported, and the carveout has not been updated since 1976 — “a time before the internet, let alone the concept of generative AI training sets.”
AD