人工智能（AI）产业高速发展，伴生了多种社会伦理风险，且相关后果已经逐步显现；让人类在充分安全的条件下享用智能技术带来的有利收益，成为AI 监管过程的重要任务。本文从AI 伦理焦点问题着手，对智能技术的机器权利与人权分配、社会信任危机、数据及算法安全性、确权与责任归属等已经出现的风险问题进行细致梳理；总结了国内外AI 产业的发展路径和政策现状，据此多维度开展我国AI 领域伦理监管的需求分析。研究认为，应根据AI 技术进展，分阶段延伸AI 治理的监管范围；建立伦理、法律、政策相结合的多维度监管框架，提供相对充分的社会言论空间，确保各利益相关方参与AI 安全公共讨论；有序完善多层级的科技伦理监管组织机构，积极参与AI 国际规则制定，为我国AI 产业的全球化发展机遇创造良好条件。

The rapid development of artificial intelligence (AI) industry is accompanied by various social ethical risks, and the consequences of these risks have gradually emerged. Allowing humans to securely enjoy the major benefits of AI technology has become an important task of AI supervision. In this article, we first analyze the focus issues of AI ethics, including the distribution of machine and human rights, social trust crisis of AI, security of data and algorithms, as well as confirmation of rights and attribution of liability. Subsequently, we summarize the development paths and policy status of AI industry in China and abroad, and analyze the demand for AI ethical supervision in China from multiple dimensions. Furthermore, we suggest that the supervision scope of AI should be extended in a staged manner based on the technical progress of AI. To create better global development opportunities for China’s AI industry, a multi-dimensional supervision framework that combines ethics, law, and policy should be established, a sufficient social discussion space should be provided for all stakeholders to participate in public discussion on AI security, scientific and technological ethics supervision organizations of multiple levels should be improved in an orderly manner, and China should actively participate in the formulation of AI international rules.