信息资源管理学报 ›› 2023, Vol. 13 ›› Issue (2): 30-41.doi: 10.13365/j.jirm.2023.02.030

• 研究论文 • 上一篇    下一篇

欧盟数据与算法安全治理:特征与启示

崔文波1 张涛1 马海群2 蔡庆平1   

  1. 1.黑龙江大学信息管理学院,哈尔滨,150080; 
    2. 黑龙江大学信息资源管理研究中心,哈尔滨,150080
  • 出版日期:2023-03-26 发布日期:2023-04-20
  • 作者简介:崔文波,硕士生,研究方向为数据分析与知识发现;张涛,博士生,副教授,硕士生导师,研究方向为政策文本计算与数据分析,Email:zhangtao@hlju.edu.cn;马海群,博士,教授,博士生导师,研究方向为信息政策与法律;蔡庆平(通讯作者),博士,讲师,研究方向为数据挖掘与分析,Email:caiqingping@hlju.edu.cn。
  • 基金资助:
    本文系国家社科基金重大项目“面向数字化发展的公共数据开放利用体系与能力建设研究”(21&ZD336)的研究成果之一。

Data and Algorithm Security Governance in the European Union: Characteristics and Implications

Cui Wenbo1 Zhang Tao1 Ma Haiqun2 Cai Qingping1   

  1. 1.School of Information Management, Heilongjiang University, Harbin 150080; 
    2. Research Center of Information Resource Management, Heilongjiang University, Harbin 150080
  • Online:2023-03-26 Published:2023-04-20

摘要: 研究全球典型区域数据与算法安全治理结构与特征,能为我国数据与算法安全治理提供参考借鉴。本文以数据与算法安全治理为对象,调研欧盟实现数字主权与技术主权治理目标的具体形式发现,在治理内容上,以《通用数据保护条例》(GDPR)为核心的欧盟数据与算法安全治理内容框架包括隐私和数据保护、数据流动安全、平台算法安全、人工智能伦理四部分;在治理主体上,以欧洲数据保护监管局等为核心,形成“立法-执行-服务”为一体的多方协同治理机构。欧盟数据与算法安全治理具有“重视个人隐私数据安全、建立数据跨境流动安全规则、形成算法嵌套数据的协同安全治理模式、以人为中心的人工智能伦理与治理”的典型特征。从欧盟数据与算法安全治理特征中得到启示,并针对我国数据与算法安全治理现状提出三点建议:完善数据与算法安全协同治理机制、建立主体间数据流动制度及中介机构评估机制、建立人工智能伦理监管与创新发展的平衡机制。

关键词: 数据治理, 算法治理, 人工智能, 算法安全, 欧盟, 数字技术, 技术主权

Abstract: Research on the structure and characteristics of data and algorithm security governance in typical global regions will provide a reference for data and algorithm security governance in China. The paper takes data and algorithm security as the governance object to investigate EU’s realization of the governance goal on digital sovereignty and technological sovereignty. In terms of governance content, the framework of the EU data and algorithm security governance with the core of the General Data Protection Regulations (GDPR) consists of four parts: privacy and data protection, data flow security, platform algorithm security, and artificial intelligence ethics. In terms of governance subjects, with the European Data Protection Regulatory Authority and other institutions as the core, a multi-party collaborative governance subject integrating "legislation-enforcement-service" is formed. The EU data and algorithm security governance has the typical characteristics of emphasizing personal privacy data security, establishing cross-border data flow security rules, forming a collaborative security governance model for algorithm nested data, and a human-centered AI ethics and governance. Based on the characteristics of data and algorithm security governance in the European Union, three suggestions are proposed for the current situation of data and algorithm security governance in China: to improve the collaborative governance mechanism of data and algorithm security, to establish a data flow system between subjects and an evaluation mechanism for intermediary agencies, as well as a balance mechanism between AI ethical supervision and innovative development.

Key words: Data governance, Algorithmic governance, Artificial intelligence, Algorithmic security, European Union, Digital sovereignty, Technological sovereignty

中图分类号: