论我国人工智能领域包容审慎监管的法治维度  

On the Legalization of Inclusive and Prudent Regulation in the Field of Artificial Intelligence

作　　者：黄锫[1] Huang Pei

机构地区：[1]同济大学法学院

出　　处：《财经法学》2025年第2期94-109,共16页Law and Economy

基　　金：国家社会科学基金一般项目“人工智能大模型包容审慎监管的法治路径研究”(24BFX031)的阶段性研究成果。

摘　　要：包容审慎监管已逐步成为我国人工智能领域的主导性监管理念。人工智能技术具有的“破坏性创新”特征决定了包容审慎监管的法治理念内核。即“包容性监管”主要是为了呵护人工智能的创新性,促进我国人工智能技术与产业的快速发展,适用于人工智能发展中遭遇的法律规则突破与法律规则空白两种情形。“审慎性监管”主要是为了应对人工智能的破坏性,防范人工智能发展中对人的权益侵害的伴生风险,适用于人工智能导致的权益侵害广度扩张与深度拓展两种情形。实现包容性监管的法治路径主要包括:通过法律解释逸脱法律规则的适用范围、灵活运用从轻/减轻/不予处罚规则、设置法定观察期等。实现审慎性监管的法治路径主要包括:将监管沙盒制度与改革试验区模式结合运用、依法划定权益保护的安全红线、设定必要的从重处罚规则等。这些研究结论都可以为我国人工智能法的立法提供参考。Inclusive and prudent regulation(IPR)has gradually become the dominant regulatory concept in the field of artificial intelligence(AI)in China.The“disruptive innovation”feature of AI technology determines the core concept of IPR.That is,“inclusive regulation”is mainly to protect the innovation of AI and promote the rapid development of AI technology and industry in China.It is applicable to the two situations where AI breaks through legal rules and where there are legal rule blanks.“Prudent regulation”is mainly to deal with the destructiveness of AI and prevent the concomitant risks of infringement on the rights and interests of human beings as the subjects in the development of AI.It is applicable to the two situations where the breadth and depth of rights and interests infringement caused by AI expand.The legal paths to achieve inclusive regulation mainly include:evading the application scope of existing legal rules through legal interpretation,flexibly applying the rules of“lighter/mitigated punishment”or“no punishment”,and setting up a“statutory observation period”,etc.The legal paths to achieve prudent regulation mainly include:combining the regulatory sandbox system with the model of reform pilot areas,legally delimiting the safety red line for rights and interests protection,and setting necessary rules of heavier punishment,etc.All these can provide references for the legislative activities of the“AI Law”in China.

关 键 词：人工智能立法 破坏性创新 包容审慎监管 包容性监管 审慎性监管 

分 类 号：D923.41[政治法律—民商法学] D922.17[政治法律—法学]