Judge Favors Anthropic Training Its AI on Copyrighted Books
In the latest twist of the AI and copyright violation saga, a California federal judge has ruled that Anthropic training its LLM Claude on copyrighted books was “fair use and “transformative” — “like any reader aspiring to be a writer.” The part of the lawsuit, brought by three authors, that focuses on Anthropic maintaining a central library of some 7 million pirated books, will proceed to trial. That claim could still result in damages.