Well, FWIW, LLMs are specified to infer and fill in the blanks of books. It makes the headlines now and again that publishers put AI companies on the hook for unauthorized use, The New Yorker included.