AI科技伦理规范文件公开征求意见，三类活动需开展专家复核

Core Viewpoint - A regulatory document for ethical governance of artificial intelligence (AI) activities is being drafted by multiple departments, with public consultation open until September 22, 2025 [2]. Group 1: Regulatory Framework - The draft document, titled "Management Service Measures for AI Technology Ethics (Trial) (Public Consultation Draft)," aims to detail and implement previous guidelines on strengthening technology ethics governance in the AI sector [2]. - The document emphasizes ethical requirements for AI activities, enhances standard construction, and provides support measures for small and medium-sized enterprises (SMEs) to improve their risk prevention capabilities [2]. Group 2: Implementation and Oversight - Institutions involved in AI activities, such as universities, research institutions, healthcare organizations, and enterprises, are designated as responsible entities for managing AI technology ethics [3]. - It is recommended that qualified organizations establish AI ethics committees, while local authorities may set up specialized AI ethics service centers [3]. Group 3: Review and Compliance - The draft outlines key review focuses for AI ethics, including fairness, controllability, transparency, and accountability [3]. - A dynamic list of AI activities requiring expert review will be established, initially including systems that significantly impact human behavior, algorithms with social mobilization capabilities, and automated decision systems with safety risks [3][4]. Group 4: Responsibilities and Penalties - The Ministry of Science and Technology will oversee national AI ethics regulation, while the Ministry of Industry and Information Technology will coordinate ethical governance efforts [4]. - Violations of the regulations will be subject to penalties in accordance with the "Science and Technology Progress Law" and other relevant laws [4].