摘要
全球贸易治理正面临人工智能时代的范式挑战。以欧盟《人工智能法案》为代表的预防性监管模式,与基于科学实证主义的WTO规则体系存在深刻冲突。AI的黑箱特性及训练过程监管要求挑战WTO核心原则。为应对挑战,WTO亟需内生性制度调适。解决方案应采取双重路径:一是在规则层面构建价值导向分类与NPR-PPMs合法性对等审查框架;二是在制度运行层面以程序正当弥补科学证据,并推动以“实质等同性”为核心的监管互认机制。该方案将审查重点从微观技术转向宏观治理实效,在兼顾成员监管自主权与维护多边贸易自由化间寻求平衡。
Global trade governance faces a paradigm challenge in the age of Artificial Intelligence.The precautionary regulatory model,exemplified by the European Union's Artificial Intelligence Act,generates a profound conflict with the rules-based system of the WTO,which is founded on scientific positivism.The"black box"nature of AI systems and regulatory demands concerning the algorithm training process challenge core WTO principles.To meet this structural challenge,the WTO urgently requires endogenous institutional adjustment.The solution should adopt a"dual-path approach":at the level of the rules,constructing a value-oriented classification and a framework for the legitimacy equivalence review of Non-Product-Related Process and Production Methods;at the institutional implementation level:utilizing procedural due process to compensate for the lack of scientific evidence,and advocating for a regulatory mutual recognition mechanism centered on"substantive equivalence".The proposed scheme aims to shift the focus from micro technology to macro-governance effectiveness,thereby achieving a balance between maintaining Member States'rIegulatory sovereignty and safeguarding multilateral trade liberalization.
作者
李萍
Li Ping(International Law and Foreign-related Rule of Law Innovation Center,Social Science Domain,Beijing Institute of Technology,Zhuhai Guangdong,519088)
出处
《政法论丛》
北大核心
2026年第2期107-122,共16页
Journal of Political Science and Law
基金
国家社会科学基金重大项目“粤港澳大湾区建设国际科创中心知识产权情报保障研究”(21&ZD339)
广东省教育厅特色创新项目“技术标准化背景下标准必要专利(SEP)纠纷解决对策研究”(2024WTSCX184)的阶段性成果
关键词
人工智能法案
WTO改革
货物-服务分类
非产品相关生产过程
预防原则
Artificial Intelligence Act
WTO reform
goods-services classification
non-product related process and production methods
precautionary principle
