A federal judge granted final approval to a $1.5 billion settlement resolving claims that Anthropic used unauthorized books to train its AI models. The agreement provides compensation to authors and publishers whose works were included in the company's digital library.