生成式人工智能虚假信息治理的新挑战及应对策略——基于敏捷治理的视角
New Challenges and Countermeasures of False Information Governance in Generative Artificial Intelligence——Based on the Agile Governance Perspective
何宇华 1李霞2
作者信息
- 1. 浙江警察学院
- 2. 中国社会科学院法学研究所宪法与行政法研究室
- 折叠
摘要
生成式人工智能的基础数据来源复杂,数据的真实性与合法性存在极大不确定性,由此出现的虚假信息在技术逻辑上具有必然性和无限复制性;在传播规律上具有瞬时性和隐蔽性;在危害后果上具有复杂性和不可控性.现有多中心协同论、二元治理结构论和层级治理论分别从不同的视角提出治理策略,但过于关注协同合作本身却忽视了治理主体的协同能力和响应效率;过度依赖政府与市场的互补却遮蔽了二者面临新技术时容易懈怠的共性;盲目信任分层治理极易陷入技治主义的困境.对此,应转向敏捷治理下以政府主导强化多元主体间的协同效能、提升治理结构的适应性、破除层级障碍的治理新思路.以虚假信息治理的法治化和技术化为基本方向,保持制度的原则性要求和技术工具价值的协同,实施虚假信息等级评估制度,构建治理主体高效协作机制、快捷响应机制和容错纠错机制,搭建生成式人工智能虚假信息敏捷治理新模式的基本框架.
Abstract
The basic data sources of generative artificial intelligence are complex and extensive,and the authenticity and legitimacy of the data are greatly uncertain.False information is inevitable and subject to infinite reproduction in the tech-nical logic.It is propagated instantly but it is also concealed and complex and capable of uncontrollable consequences.The multi-center synergy theory,dual governance structure theory and hierarchical governance theory suggest governance strat-egies from different perspectives.But theyfocus too much on collaborative cooperation and ignore the collaborative abilities and response efficiencies of the governance subjects.They also rely on the complementarity of government and the market to explain their common inclination to become slack in the face of new technologies.A blind trust in hierarchical govern-ance easily leads to the dilemma of technocracy.We should turn instead to a new governance idea under agile governance.We should strengthen collaborative efficiency among multiple subjects under the government's leadership,improve the a-daptability of governance structure through organizational flexibility,and break down hierarchical barriers through penetrat-ing governance.If legalization and technicalization of false information governance are the basic direction,then the princi-ple requirements of the system and the synergy of the value of technical tools are maintained;the false information grade e-valuation system is implemented;and,the efficient cooperation mechanism,quick response mechanism,and fault toler-ance and error correction mechanism of the governance bodies will be constructed.Ultimately,this will lead to the con-struction of the basic framework of a new model of agile governance of false information in generating artificial intelligence.
关键词
生成式人工智能/虚假信息/敏捷治理Key words
generative artificial intelligence/false information/agile governance引用本文复制引用
基金项目
2024年度浙江省哲学社会科学规划课题(24NDJC164YB)
浙江省教育科学规划2022年度一般规划课题(高校)(2022SCG320)
中国社会科学院-上海市人民政府上海研究院交办项目(2023jb023)
出版年
2024
NETL
NSTL
万方数据