基于Post-ELSI的人工智能政策国际比较与经验镜鉴  

International Comparison and Experience Learning of Artificial Intelligence Policy from the Perspective of Post-ELSI

作　　者：张成岗 阿柔娜 ZHANG Cheng-gang;A Rou-na(Department of Sociology,Tsinghua University,Beijing 100084,China)

机构地区：[1]清华大学社会学系,北京100084

出　　处：《江苏行政学院学报》2023年第3期69-76,共8页The Journal of Jiangsu Administration Institute

基　　金：国家社科基金重大研究专项“一带一路’建设面临的主要风险及应对研究”(18VDL015);贵阳市南明区“社会治理现代化专题研究”(2022002068)的阶段性成果。

摘　　要：人工智能技术发展带来的社会风险日益受到关注,各国和欧盟均致力于通过制定和实施人工智能政策形塑人工智能技术,Post-ELSI为人工智能政策的评估提供对伦理、法律和社会因素更全面和细致分析的新思路。本文基于Post-ELSI理论框架,运用NVivo12 plus质性资料的数据分析工具对2016-2020年中国、美国、韩国、日本、欧盟等发布的29份人工智能政策,从伦理(安全、隐私、责任等)、法律(确权、监管、算法等)、社会(公正、平等、治理等)三方面进行评估分析。研究发现,在Post-ELSI框架下,中国、美国、韩国、日本和欧盟的人工智能政策具有很大同质性,均将“技术精英”与“技术难民”、“人机协作”与“机器换人”、“对症下药”与“被人工智能评估的人”视为关切利益,但在伦理危机、负责任监管和社会福祉等方面的关注程度与政策措施存在较大差异性。通过镜鉴国际经验,可以为制定具有中国特色的“负责任的”人工智能政策提供有力支撑。In the context of the increasingly prominent social risks brought about by the development of artificial intel-ligence technology.Post-ELSI provides a new way of evaluation for AI policies with a more comprehensive and detailed analysis of ethical,legal and social factors.Based on the theoretical framework of Post-ELSI,this paper uses the data anal-ysis tool of NVivo12 Plus qualitative data to analyze 29 AI policies issued by China,the United States,the Republic of Ko-rea,Japan,the European Union and other countries and regions from 2016 to 2020.The data had been evaluated and ana-lyzed from three aspects:ethics,law and society.The study found that under the Post-ELSI framework,the AI policies of China,the United States,the Republic of Korea,Japan and the European Union show great trend of homogeneity,and all regard"technical elites"and"technical refugees","human-computer collaboration"and"machine substitution",and"symptom-specific medicine"and"people being evaluated by Al"as concerns.However,there are great dfferences in the degree of concern and policy measures in ethical crisis,accountability,supervision,and social welfare.Learning from international experience can provide strong support for making"responsible"AI policies with Chinese characteristics.

关 键 词：人工智能 伦理 法律 社会 国际比较 

分 类 号：C934[经济管理—管理学]