Path Choice for Artificial Intelligence Regulation:Paradigms,Controversies and Impacts of the EU's Artificial Intelligence Act
China's AI legislation has already been put on the agenda,but there is yet no consensus on the regulatory path to be followed.The previous attempts in the other jurisdictions can thus be used for reference.The EU Artificial Intelligence Act is an important step in Europe's digital legislation and the cornerstone of the EU's AI strategy,which follows a horizontal legislative model and opts for a"risk-based"approach,categorizing AI systems into four risk levels with corresponding regulatory measures:banned AI systems with unacceptable risks,high-risk AI systems subject to specific rules,other AI systems,and Generic AI models.The is-sues of"real-time"remote biometric identification in public spaces and the regulation of general-purpose AI models are highly controversial.While at the same time,the EU's Al Law faces predicaments in relation to standardization,increased costs for market players and high implementation costs.The EU has chosen a different regulatory path for AI regulation from that taken by the US or the UK.China should choose a regulatory path combining both horizontal and vertical models which is based on its own AI strategy.
万方数据
