The proposed rules would be among the toughest global AI regulations if passed. Minors and elderly users would be required to register a guardian to use AI services, who’d be notified if topics like suicide come up, and chatbots would be banned from emotional manipulation and promoting violence, crime, or self-harm.