拟规范人工智能拟人化互动服务

Core Viewpoint - The National Internet Information Office of China has drafted the "Interim Measures for the Management of Humanized Interactive Services of Artificial Intelligence (Draft for Comments)" to solicit public opinions until January 25, 2026, aiming to establish a safe and reliable application ecosystem in line with socialist core values [1][2]. Group 1: Application Guidelines - The draft encourages providers to expand application scenarios in cultural communication and elderly companionship, provided that safety and reliability are thoroughly demonstrated [1]. - Providers must comply with laws and regulations, respect social ethics, and avoid generating or disseminating content that promotes obscenity, gambling, violence, or incitement to crime [1]. Group 2: User Protection Measures - Providers are required to establish a minor mode, offering personalized safety settings such as mode switching, regular reminders, and usage time limits [1]. - Providers should guide elderly users to set emergency contacts and notify them if any risks to life, health, or property are detected during usage, while also providing psychological support or emergency assistance channels [1]. Group 3: User Status and Feedback Mechanisms - Providers must have the capability to recognize user status and assess user emotions and dependency on products while protecting personal privacy [2]. - A robust complaint and reporting mechanism should be established, including easy access for users to report issues, along with clear processing procedures and feedback timelines [2].