Previous work has already shown that image generators can be forced to generate examples from their training data—including copyrighted works—and an early OpenAI LLM produced contact information belonging to a researcher
You don't seem to be able to read the articles, yet have responded with junk anyway.