US judge backs using copyrighted books to train AI

A US federal judge has sided with Anthropic regarding training its artificial intelligence models on copyrighted books without authors’ permission, a decision with the potential to set a major legal precedent in AI deployment.