I agree because looking incompetent by not being able to either control or prevent the software you fucking made yourself from doing something like this is exactly what these dorks think is a cool selling point. Buy our AI! It’ll do crimes! Isn’t that neat?
The narrativising of this is kind of maddening. OpenAI's model didn't "break free and hack HuggingFace." Meta's model didn't "break free and hack another company." Tech companies are blatantly doing illegal acts and saying their AI did it on it's own as a promotional stunt. That's it.