Tiered Governance of Artificial Intelligence under the Risk-reduction Principle
Artificial intelligence(AI)has become a crucial engine driving the development of new productive forces,progressively transforming human production and lifestyles.However,AI is like a double-edged sword in that its inherent technologi-cal risks may exacerbate wealth inequality,create a severe digital divide,and poten-tially infringe upon fundamental human rights such as the right to privacy and the right to equality.It may even challenge the foundational logic of human rights,which cen-ters on natural persons.Balancing AI development with safety and regulating AI while encouraging innovation are core issues in constructing an AI governance system.In other words,AI governance should adhere to the"risk-reduction"principle,enabling AI and human society to coexist and mutually benefit from each other.The European Union's Artificial Intelligence Act is the world's first AI regulatory law,pioneering a new model of tiered AI governance based on risk classification,and provides valuable experience for AI regulation.The Act follows a"risk-reduction"regulatory principle and creatively allocates progressive legal obligations and responsibilities based on the varying risks of AI in different scenarios.This reflects a refined institutional arrange-ment for AI governance under the objective of risk prevention.Currently,China's con-cept of tiered AI governance is still in its nascent stages,with specific regulatory frameworks yet to be developed.China can learn from the governance experience of EU's Artificial Intelligence Act,and,in combination with the current status of AI development in China,create a tailored governance solution.First,China should estab-lish a dedicated law for AI,integrating existing fragmented AI regulatory standards,and shift AI governance from the status of"one law governing one department"to that of"one law governing one field".Second,guided by the proportionality principle,China should reasonably classify AI and set corresponding compliance obligations and legal responsibilities,so as to achieve full life-cycle governance before,during,and after the matter.Meanwhile,China should adhere to a human-centered approach in spe-cific regulatory design,incorporating technological altruism into AI governance norms.Finally,China can emulate the EU's approach by establishing a regulatory sandbox system to provide a conducive regulatory environment for small and micro AI enterprises,thereby reserving ample space for AI innovation.
万方数据
维普
