摘要
[研究目的]大模型技术引发了情报技术革命,使数据跨境隐性国家安全风险成为不逊色于显性风险的重要类型,亟需治理。[研究方法]通过文献分析、案例分析、法规梳理等方法对大模型时代数据跨境隐性国家安全风险的治理现状、治理困局等进行了研究,发现当下以事前预防为主要特征的科层化治理模式与隐性数据安全风险特性相冲突,面临治理对象失焦、治理重心错位和治理工具针对性差的困局。[研究结果/结论]敏捷治理因应新兴风险不确定的特性而被提出,与数据跨境外部政治环境和隐性风险演化的双重复杂性相适应,应予以引入。据此,治理思路上应推动风险预判从静态内容分级到动态行为评估、治理重点从内部防御到跨境协同和治理方法从静态合规到动态干预升级转变。为贯彻敏捷治理的思路,应从出境前风险过滤机制、跨境后动态反制及多元主体协同共治方面予以落实。
[Research purpose]Big model technology has triggered a revolution in intelligence technology,making cross-border data implicit national security risks no less important than explicit risks that need to be urgently managed.[Research method]The governance status quo and governance predicament of cross-border data implicit national security risks in the era of big models are studied through literature analysis,case analysis,and regulation combing,etc.The study's findings indicate that China's current sectionalized governance framework,which is primarily characterized by ex-ante prevention,is incompatible with the characteristics of implicit data security risks.It is also faced with the predicament of out-of-focus governance objects,misplaced governance focus,and poorly targeted governance tools.[Research result/conclusion]Agile governance,proposed in response to the uncertain character of emerging risks,is compatible with the dual complexity of cross-border data external political environment and implicit risk evolution,and should be introduced.Accordingly,the governance mindset should undergo an agile transformation,promoting risk prediction from static content grading to dynamic behavioral assessment,governance focus from internal defense to cross-border collaboration,and governance methodology from static compliance to dynamic intervention.In order to implement the idea of agile governance,it should be implemented in terms of pre-departure risk-filtering mechanisms,post-cross-border dynamic countermeasures,and collaborative governance by multiple actors.
作者
杨胜刚
李川
Yang Shenggang;Li Chuan(Southeast University,Nanjing 211189)
出处
《情报杂志》
北大核心
2025年第8期51-58,共8页
Journal of Intelligence
基金
国家社会科学基金青年项目“行刑共治视域下滥用数据要素信用惩戒机制研究”(编号:24CFX099)研究成果。
