生成式AI“未保”怎么做？专家：建保护模式，平衡管控体验

Core Viewpoint - Generative AI is rapidly integrating into the digital lives of minors, providing learning assistance and social companionship, but it also raises significant concerns regarding the potential harm to minors, including exposure to inappropriate content and over-reliance on AI tools [1][2][4]. Group 1: Current Usage and Statistics - According to the 55th CNNIC report, 21.1% of internet users aged 6-19 are utilizing generative AI products, with many minors engaging in activities such as using AI for problem-solving, writing, and social chatting [2]. - The South Data Research Institute has observed that applications like "smart Q&A," "creative generation," and "virtual social interaction" are among the most commonly used generative AI services by minors [4]. Group 2: Guidelines and Frameworks - The "Guidelines for Providing Generative AI Services to Minors" was released on June 10, establishing a comprehensive safety management framework covering the entire lifecycle from training data to service operation [2][3]. - The guidelines emphasize the need for a protective framework that includes measures to prevent over-reliance, exposure to harmful information, and privacy breaches [3][4]. Group 3: Expert Recommendations - Experts suggest that generative AI applications should incorporate a "minor mode" that includes features such as age-appropriate content filtering, identity verification, and parental controls to enhance safety [5][6]. - Recommendations for the design of the "minor mode" include easy-to-use features, age-based content recommendations, and mechanisms for reporting harmful content [6][7]. Group 4: Regulatory and Compliance Measures - Existing regulations, such as the "Interim Measures for the Management of Generative AI Services," outline the responsibilities of service providers to protect minors from excessive reliance on AI [3]. - The establishment of a comprehensive evaluation index system for the protection and development of minors in the context of generative AI has been initiated, focusing on areas such as information security and emergency response mechanisms [3].