
Encyclopedia Britannica and Merriam-Webster say that OpenAI violated the copyright of almost 100,000 articles by using them for LLM training.
Source

Close-up of a yellow-orange pen poised over printed text, with the word copyright clearly visible.
Posted inblog
