AI companies have faced several copyright lawsuits over the ethics of using published literature to train their models.

Anthropic destroying books to train Claude was ruled legal. Here is the rule that made it work, explained simply.