Encyclopedia Britannica and Merriam-Webster have sued OpenAI, alleging ChatGPT trained on and reproduces nearly 100,000 of their copyrighted articles.