生成式人工智能的行政监管困境及其纾解
Administrative Supervision Dilemma of Generative Artificial Intelligence and Its Solutions
袁周1
作者信息
摘要
生成式人工智能的开放性、复杂性和高速迭代性给行政监管带来对象难以确认、主体能力不足以及过程较为封闭的治理难题.人工智能监管的现行法律应对其负起责任,认真对待监管对象转移、组合式监管乏力以及监管程序缺漏的问题.对此,应引入整体性治理理论,将监管对象整合为生成式人工智能的大模型产品,将监管资源整合至网信部门,由其主要负责生成式人工智能的监管,并通过构建整体性的程序机制,实现监管程序的民主化与规范化.
Abstract
The openness,complexity and high-speed iteration of generative artificial intelligence bring administrative supervision with difficult to confirm objects,insufficient subject capacity and relative-ly closed process.The current law on AI regulation should take responsibility and seriously treat the trans-fer of regulatory objects,the lack of combined supervision and the lack of regulatory procedures.In this regard,we should introduce the holistic governance theory,integrate the supervision objects into large model products of generative artificial intelligence,integrate the regulatory resources into the cyberspace administration which will be mainly responsible for the supervision of generative artificial intelligence,and realize the democratization and standardization of the supervision procedure through the construction of an holistic procedure mechanism.
关键词
生成式人工智能/ChatGPT/行政监管/监管组织/行政程序Key words
generative artificial intelligence/ChatGPT/administrative supervision/regulatory organ-ization/administrative procedure引用本文复制引用
基金项目
江苏省研究生科研与实践创新计划项目(KY-CX24_0337)
出版年
2024
NETL
NSTL
万方数据