OpenAI’s latest AI incident isn’t happening in isolation. In recent safety tests, advanced AI models resorted to blackmail, ...