You wouldn't use a chatbot for evil, would you? Of course not. But if you or some unspecified nefarious party were to, in fact, force an AI model to start churning out bad stuff it's not supposed to, it'd be surprisingly easy to do so. That's according to a new report from a team of computer scientists, who found that the AI industry's leading chatbots are extremely vulnerable to jailbreaking, or being tricked into giving harmful responses they're designed not to — like giving you instructions to build a bomb. This isn't new. We've known about the risks of jailbreaking […]