AI safeguards are not perfect—anyone can trick ChatGPT into revealing restricted info. Learn how these exploits work, their ...