反身法视野下生成式人工智能的耦合型监管

Coupling Supervision of Generative Artificial Intelligence from the Perspective of Reflexive Law

生成式人工智能的快速发展带来了风险控制和监管难题。从社会系统论角度分析,欧盟基于风险的监管模式和美国基于场景的监管模式,均未能实现法律子系统和生成式人工智能子系统之间的有效互动,从而面临多重现实困境。反身法提倡“受规整的自治”,通过反思机制为其他社会子系统的“自我规制”提供制度保障,从而建构“生态自组织”的监管形态,对突破生成式人工智能的监管困境具有启发意义。立足反身法框架,借助战略生态位管理模型,厘清生成式人工智能“技术生态位—市场生态位—范式生态位”的内在演化逻辑,有助于探索和塑造包容审慎与敏捷高效相结合的、融外部监管与自我合规于一体的耦合型监管模式。

The rapid development of generative AI has brought about challenges of risk control and regulation.From the perspective of social system theory,both the risk-based regulatory model of the EU and the scenario-based regulatory model of the US fail to achieve effective interaction between the legal subsystem and the generative artificial intelligence subsystem,thus facing multiple realistic dilemmas.Reflexive law advocates"regulated autonomy"and provides institutional guarantee for the"self-regulation"of other social subsystems through the reflection mechanism so as to construct the regulatory form of"ecological self-organization",which has enlightening significance for breaking through the regulatory dilemma of generative artificial intelligence.Based on the reflexive law framework and with the help of the strategic niche management model(SNM),the internal evolutionary logic of the generative artificial intelligence"technology niche-market niche-paradigm niche"is clarified,which is helpful to explore and shape the coupling regulatory model that combines inclusive prudence with agility and efficiency,and integrates external supervision and self-compliance.