Anthropic must pay thousands of impacted authors.

The dispute is one of several in which copyright holders have sued AI companies over how they have trained their large language models (LLMs).

The final approval settles one case, but it doesn't resolve the broader issue of using copyrighted works to train AI models.