中央网信办于永河：AI标识制度获公众广泛认可，将健全完善

Core Viewpoint - The implementation of the AI-generated synthetic content identification system since September 1, 2025, has gained widespread public recognition, establishing a comprehensive and innovative governance framework for AI safety in China [2] Group 1: AI Governance System - The AI identification system reflects a systematic approach to AI governance, building on previous regulations related to content generation and dissemination [2][3] - The system consists of a three-tier structure: a regulatory document, a mandatory national standard, and a set of practical guidelines, forming a "1+1+N" system [3] Group 2: Content Generation and Distribution - The AI identification system governs the entire content generation and distribution chain, requiring content creation platforms to embed identifiers and distribution platforms to verify and update these identifiers [3] - This collaborative approach addresses key questions regarding the origin and nature of generated content [3] Group 3: Innovation in Content Identification - The system introduces innovative explicit identifiers for text and audio content, such as corner marks for text and Morse code for audio, minimizing user disruption while enhancing identification [4][5] - Over 1.5 trillion pieces of text, images, and audio-visual content have been tagged with AI identifiers by major platforms, with over 1 billion files provided to users [5] Group 4: Public Awareness and Acceptance - A survey indicated that 76.4% of respondents noticed increased content identification on social media and news platforms, with 60% believing it aids in distinguishing AI-generated content [5] - The public has developed an awareness of "learning, recognizing, and using" identifiers [5] Group 5: Future Directions - Continuous efforts are needed to enhance the AI identification system, including experience accumulation, regulatory enforcement, public education, and fostering international consensus on AI governance [6]