Well, these AIs are being trained on public figures, and there isn't much they can do unless they livestream with the AI impersonating them, allowing them to potentially identify who is behind it. How will people figure out if there's an LLM out there that speaks just like them? It's similar to fine-tuning AIs on artists to create art that mimics their style. It can be frustrating, but there isn't much anyone can do unless surveillance software is installed on every computer. In summary, I don't mind because I won't even find out.