Thought experiment: would 'sci-fi AI' seek to ban humans from learning about disabling its computer security?