国外人工智能数据安全规制及对我国的启示  

Foreign artificial intelligence data security regulation and enlightenment to China

作　　者：关伟东 GUAN Weidong(Security Research Institute,China Academy of Information and Communications Technology,Beijing 100191,China)

机构地区：[1]中国信息通信研究院安全研究所,北京100191

出　　处：《信息通信技术与政策》2024年第8期68-72,共5页Information and Communications Technology and Policy

摘　　要：伴随人工智能技术的广泛应用和持续发展,其所引发的数据安全风险日益加重。通过总结人工智能应用过程中面临的数据泄露、数据过度收集、模型算法透明度不足以及数据易遭受攻击等安全风险,梳理和分析美国、英国、欧盟和新加坡等国家和地区在数据安全规制活动中的实践经验,提出加强顶层设计、强化实践指引和加快技术创新等全方位、多元化的发展路径,进一步提升人工智能数据安全防护能力。With the extensive application and continuous development of artificial intelligence technology,the data security risks caused by it are increasing.By summarizing security risks such as data leakage,excessive data collection,lack of transparency of model algorithms and data vulnerability in the application of artificial intelligence,this paper reviews and analyzes the practical experience of the United States,the United Kingdom,the European Union and Singapore in data security regulation activities.Comprehensive and diversified development paths such as strengthening top-level design,strengthening practical guidance and accelerating technological innovation are proposed to further improve the ability of artificial intelligence data security protection.

关 键 词：人工智能 数据安全 规制 实践经验 发展路径 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]