人工智能立法的一般范畴

The General Scope of Artificial Intelligence Legislation

当前人工智能的治理已经迈入大规模立法的新阶段,但相关研究呈现出明显的部门法色彩与具象化特征,亟需透过具体问题的迷雾从整体上构建人工智能立法的一般范畴。人工智能立法应当区分智能科技、智能要素以及智能应用三个基本层次,不同层次的风险发生、表征与治理机制不同。人工智能立法需要产业促进法与风险治理法两组规范类型,产业促进法又具体包括市场法与政策法两组子类型,既要克服科技不足也要警惕科技过度,并明确作为立法内容的主体、客体以及权责关系。协同治理、多元治理以及敏捷治理应当成为人工智能立法的基本思维,同时有必要制定一份清晰的人工智能立法路线图,以便明确未来人工智能立法的整体方向与具体步骤。

At present,the governance of artificial intelligence has entered a new stage of large-scale legislation,however,the research on AI governance has presented a piecemeal and fragmented feature,so it is urgent to pierce the fog of specific issues and construct the general scope of artificial intelligence legislation.Artificial intelligence legislation should distinguish three basic levels of intelligent technology,elements,and application,which correspond to different risk probability,characteristic and governance mechanism.And it should include two types of norms:industry promotion law and risk management law,with industry promotion law including two sub-types of market law and policy law.In the legislative process,it is important to overcome technological deficiency on one hand,and prevent excessive technology on the other hand.Besides,the subject,object and rights and obligations should be clearly laid down.We should take the approach of collaborative,multi-party and efficient governance when legislating on AI.Meanwhile,it is necessary to make a clear road map,so as to define the overall direction and specific steps of artificial intelligence legislation in the future.