阅江学刊2025，Vol.17Issue(5)：119-133,15.
颠覆与重构:生成式人工智能的信息风险源发与治理进路
Disruption and Reconstruction:Information Risk Origins and Governance Approaches in Generative AI
摘要
Abstract
Generative AI technologies,exemplified by ChatGPT,Claude,Sora,and DeepSeek,have rapidly evolved into a new type of infrastructure.Their widespread adoption has brought disruptive impacts to complex social information dissemination systems.Scholars from diverse disciplines have engaged in vigorous discussions regarding the origins of information risks and governance approaches for generative AI.The paper analyzes the technical,content,and regulatory risks posed by generative AI and proposes corresponding governance strategies and mechanisms.Generative AI communication has fundamentally transformed human information dissemination paradigms,rendering traditional governance models increasingly ineffective.China's current AI governance landscape faces challenges including excessive content control,fragmented information management systems,diverse governing entities,and ambiguous rights and responsibilities.Under the framework of overall national security,key issues requiring resolution in generative AI governance include balancing development with security,ensuring content production quality and compliance,and implementing effective risk management.关键词
生成式人工智能/信息风险/智能治理/智能伦理引用本文复制引用
陈力双,沈天健,张文祥..颠覆与重构:生成式人工智能的信息风险源发与治理进路[J].阅江学刊,2025,17(5):119-133,15.基金项目
国家社会科学基金重大项目"人工智能时代的新闻伦理与法规"(18ZDA308). （18ZDA308）
