摘要
[目的/意义]探究生成式人工智能对全球安全治理的挑战,解析联合国及其专门机构的治理政策演变,提出治理模式优化路径。[方法/过程]以联合国32份现行人工智能治理政策文本为对象,按照时间周期对政策文本进行划分,揭示国际组织人工智能治理政策布局特征与转型逻辑。采用政策文本量化研究框架,结合BERTopic主题模型,对政策文本进行时序量化分析。[结果/结论]联合国提倡以国家间协同共建为准则的协作治理模式,关注发展中国家的治理能力建设,重视技术的可持续发展与人权保障。
[Objective/Significance]Explore the challenges of generative artificial intelligence to global security governance,analyze the evolution of governance policies of the United Nations and its specialized agencies,and propose paths to optimize governance models.[Method/Process]Taking 32 current artificial intelligence governance policy texts of the United Nations as the research object,the policy texts are divided according to time periods to reveal the characteristics of policy layout and transformation logic of artificial intelligence governance by international organizations.Adopting a quantitative research framework for policy texts,combined with the BERTopic model,to conduct a time-series quantitative analysis of the policy texts.[Results/Conclusions]The United Nations advocates a collaborative governance model based on the principle of synergistic co-construction among countries,pays attention to the building of governance capacity in developing countries,and attaches importance to the sustainable development of technology and the protection of human rights.
作者
陈璐
程清君
陈方
Chen Lu;Cheng Qingjun;Chen Fang(National Science Library(Chengdu),Chinese Academy of Sciences,Chengdu 610041,China;Department of Information Resources Management,School of Economics and Management,University of Chinese Academy of Sciences,Beijing 100190,China)
出处
《科学观察》
2025年第6期51-62,共12页
Science Focus
基金
中国科学院文献情报能力建设专项“生物领域科技态势战略研判”(E4C0002301)。
关键词
人工智能
政策文本量化
联合国治理
主题建模
artificial intelligence
policy text quantifies
the United Nations governance
theme modeling
