人工智能的监管困境及其化解

The Regulatory Dilemma of Artificial Intelligence and Its Resolution

监管沙盒是源自金融科技领域的一项监管方法,在人工智能领域的运用已经得到欧盟的实践。监管沙盒作为一种良性互动型监管、实验型监管的范式,有效平衡了安全与创新,且完美体现了包容审慎和敏捷治理的监管理念,不仅为人工智能的各种风险提供缓冲区,而且还以实验性监管面对破坏性创新,有效突破传统监管模式的困境。我国应当抓紧构建中国特色人工智能监管沙盒,转变传统监管模式和理念。在具体制度构建上,需要统一监管实施主体、明确准入和退出规则、建立豁免机制。

The development and growth of artificial intelligence in society has been an undeniable fact.The development of new technology is a double-edged sword.As a product of destructive innovation,artificial intelligence brings a series of social risks and its"destructive innovation"and"uncertainty"characteristics also make the traditional regulatory model indicate certain difficulties.Regulatory sandbox,as a regulatory method in the field of financial technology,has been practiced in the EU in the field of artificial intelligence.As a paradigm of benign interactive regulation and experimental regulation,regulatory sandbox effectively balances safety and innovation and perfectly embodies the regulatory concept of inclusive prudence and agile governance,which not only provides a buffer zone for various risks of AI,but also confronts destructive innovation with experimental regulation,effectively breaking through the dilemma of traditional regulatory model.China should seize the opportunity to build an AI regulatory sandbox with Chinese characteristics and transform the traditional regulatory model and concept.In terms of specific system construction,there is need to unify the main body of regulatory implementation,clarify the rules of entry and exit and establish an exemption mechanism.