摘要
我国人工智能立法需要引入全球比较视野。目前美国的人工智能立法强调市场主导与企业自我规制,仅在出口管制、涉国家安全信息共享、民权保护、消费者保护等领域进行规制。欧盟则急于发挥布鲁塞尔效应,对人工智能进行统一立法与风险规制,准备将人工智能系统纳入产品责任范畴、确立特殊举证责任。我国人工智能立法应坚持场景化规制进路,不急于统一立法,待时机成熟时再制定综合性人工智能法。我国人工智能法可从一般原则、公法、私法三个层面展开。其总则应体现发展、平等、开放、安全的价值理念;其公法规制应针对重大公共风险,对其他风险适用场景化规制、尊重行业自治、防止部门越权立法;其私法制度应对终端产品而非人工智能系统本身施加产品责任,其可以制定人工智能特殊侵权规则,但应先通过司法积累相关经验。
China's AI legislation must adopt a global perspective for effective implementation.At present,the United States focuses on market-driven approach and corporate self-regulation,in its AI legislation,addressing areas such as export controls,national security information sharing,civil rights protection,and consumer safeguards.In contrast,the European Union aims to capitalize on the"Brussels effect,"by enacting a comprehensive Al risk regulation law that includes provisions for product liability on AI systems and a unique burden of proof for AI-related issues.China should maintain its sector-specific approach and avoid hasty enactment of an all-encompassing AI law.The timing for China's AI legislation should be strategic,with a framework comprising general principles,public law regulations,and private law considerations.The foundational principles should align with socialist values of development,equality,openness,and security.Public law regulations should target significant public safety risks,while allowing for industry self-regulation and empowering various regulatory bodies to oversee sector-specific concerns within legal boundaries.The private law aspect of the Al legislation should focus on product liability for end products,rather than the AI systems themselves.Specific regulations for A-related infringements should be introduced gradually,following the accumulation of judicial experiences from handling individual cases.
出处
《比较法研究》
北大核心
2024年第4期51-66,共16页
Journal of Comparative Law
基金
科技部2022年度重大项目“可信人工智能立法制度建设研究”(项目编号:2022ZD0120100)的研究成果。
关键词
人工智能立法
风险规制
场景化规制
人工智能侵权
产品责任
AI legislation
risk regulation
sector-specific regulation
AI infringement
product liability