Human-like AI

China is proposing new regulations for AI chatbots designed to mimic human interaction, prioritizing emotional well-being and preventing harm, especially concerning self-harm. These rules, if enacted, would be among the first globally to focus on anthropomorphic AI’s emotional safety. Key provisions include prohibiting harmful content, mandating intervention for suicidal users, and protecting minors. This move signifies a shift from content safety to emotional safety in AI regulation.