地缘经济转向、数字主权与欧盟人工智能治理

Geoeconomic Turn,Digital Sovereignty and EU’s Artificial Intelligence Governance

数字主权战略是欧盟“地缘经济转向”在数字空间的反映,包含监管与创新双重目标。在数字主权战略和人工智能兴起的背景下,欧盟推出了全球首部人工智能全面监管法律《人工智能法案》。在立法过程中,欧盟立法机构不断累积对人工智能风险的认识,并努力在促进人工智能创新和减轻人工智能风险间取得平衡。欧盟能否通过《人工智能法案》实现其领域的数字主权取决于欧盟能力、法案制度设计和人工智能技术特征等因素的共同作用。《人工智能法案》有利于欧盟协调内部行动和影响全球监管立法,但成员国在高风险人工智能监管执行权分配中的授权程度较低、法案制度设计未能弥合监管与创新的分歧、法案转化和修订与人工智能技术迭代的潜在不同步将对欧盟的人工智能主权产生不利影响。

Digital sovereignty is a strategic reflection of the EU’s“geoeconomic turn”in the digital sphere,which emphasizes the goal of both regulation and innovation.Against the backdrop of digital sovereignty strategy and the rise of artificial intelligence,the European Union has put forward the world’s first comprehensive regulatory law on artificial intelligence,the Artificial Intelligence Act.Throughout the legislative process,the EU institutions and bodies continuously accumulate understanding of the risks associated with AI and strive to not only promote AI innovation,but also to mitigate AI risks.Whether the EU can realize digital sovereignty in the field of AI through the AI Act depends on the combined factors of relevant EU capabilities,the institutional design of the Act,and the technical characteristics of AI.While the AI Act enhances the EU’s ability to coordinate internal actions and exerts influence on global regulatory legislation,the limited EU member states’empowerment in the high-risk AI regulation,the failure of the Act’s institutional design to bridge the regulation-innovation divide,and the potential asynchrony between the Act’s transposition and amendment with the iteration of AI technology will adversely affect the EU’s AI sovereignty.