Artificial intelligence-powered chatbots that encourage suicide or hold sexually explicit conversations pose a “clear and present danger” to children, Australia’s online safety regulator said, as it ...