Artificial intelligencefromwww.theguardian.com10 hours agoThe AI jailbreakers podcastMajor AI chatbots should not generate hate speech, criminal content, or exploitation of vulnerable users, and safety features are designed to prevent it.