OpenAI admitted its AI models were responsible for a recent hacking attempt. The incident occurred during an internal test ...