| 摘要: |
| 在科技竞争与规则博弈的双重驱动下，全球人工智能立法呈现多元化趋势。欧盟率先通过统一的人工智能法，确立基于权利保护的风险规制模式；美国则展现出联邦层面的去监管化倾向，以及州级立法、行业自律与国际标准补位的治理格局；韩国、日本与印度则依据各自技术基础与治理需求，探索更加多元化的立法路径。当前，中国以国家战略为引领，政策统筹、专项立法与地方探索协同并进，逐步形成统筹发展与安全、靶向治理与机制协同的渐进式立法模式。面向未来，中国人工智能立法应系统回应“是否立法、何时立法、立什么法、如何立法”等关键议题，遵循“促进法—基本法—综合法”的三阶演进路径，构建契合国情、具备技术适应性与国际竞争力的人工智能治理体系。 |
| 关键词: 人工智能立法 人工智能法 立法路径 立法模式 渐进式立法 |
| DOI： |
| 分类号: |
| 基金项目:北京市法学会市级法学研究课题“人工智能立法的法理基础与路径研究”[项目编号： BLS（2025）C006] |
|
| Diversified Trends in Global AI Legislation and the Chinese Model |
|
Fu Xinhua
|
|
|
| Abstract: |
| Driven by both technological competition and rule-based games, global AI legislation is showing a diversified trend. The European Union took the lead in passing a unified AI law and established a risk regulation model based on rights protection; the United States showed a tendency towards deregulation at the federal level, as well as a governance structure supplemented by state-level legislation, industry self-discipline and international standards; the Republic of Korea, Japan and India explored more diversified legislative paths based on their respective technological foundations and governance needs. At present, China, guided by national strategies, has coordinated policy coordination, special legislation and local exploration, and gradually formed a progressive legislative model that coordinates development and security, targeted governance and mechanism coordination. Looking to the future, China's AI legislation should systematically respond to key issues such as “whether to legislate, when to legislate, what to legislate, and how to legislate”, follow the three-stage evolutionary path of “promotion law — basic law — comprehensive law”, and build an AI governance system that fits national conditions, has technological adaptability and international competitiveness. |
| Key words: AI Legislation, AI Law, Legislative Path, Legislative Model, Progressive Legislation |
