网信办征求意见：人工智能拟人化互动服务提供者应当建立应急响应机制

Core Viewpoint - The National Internet Information Office is soliciting opinions on the "Interim Measures for the Management of Humanized Interactive Services of Artificial Intelligence (Draft for Comments)" which emphasizes the need for service providers to recognize user states and intervene in cases of extreme emotions or addiction while protecting user privacy [1] Group 1 - Service providers must have the capability to assess user emotions and dependency on products and services while ensuring user privacy [1] - Providers are required to implement preset response templates to address high-risk tendencies that threaten user health and safety, offering reassurance and encouraging help-seeking behavior [1] - An emergency response mechanism must be established to take over conversations when users express intentions of self-harm or suicide, including contacting guardians or emergency contacts [1] Group 2 - For vulnerable groups such as minors and the elderly, service providers must collect guardian and emergency contact information during the registration process [1]