A news summary of the article body: Large language models (LLMs) like ChatGPT, Gemini, and Claude train on vast text databases, often without permission from content creators. The largest settlement to date involved Anthropic, which agreed to pay $1.5 billion ($3,000 per book) to authors of 500,000 books, though a federal judge ruled the misdeed was acquiring books through piracy websites, not training on copyrighted books. In 2025, researchers found a Meta LLM had memorized and could regurgitate most of J. K. Rowling’s first Harry Potter book and George Orwell’s Nineteen Eighty-four. In March 2026, Encyclopedia Britannica and Merriam‒Webster filed a lawsuit against OpenAI, alleging it used nearly 100,000 copyrighted online articles without permission and that its LLMs violate copyright laws when generating output containing “full or partial verbatim reproductions.” The tech companies’ argument that LLM content is fresh exposes them to defamation liability, as LLMs generate and publish content. In May 2026, a German regional court issued a temporary injunction on this matter.