(Yicai) April 9 -- China has released trial Measures for AI Science and Technology Ethics Review and Service, the country's first dedicated regulations aimed at enhancing the high-quality development of the industry and preventing risks brought by innovative technologies. 

The measures, issued by the Ministry of Industry and Information Technology and nine other agencies, identified the development of human-machine fusion systems that significantly affect human subjective behavior, psychological states, or physical health, the development of algorithmic models and application systems capable of shaping public opinion or steering social consciousness, and the development of highly autonomous decision-making systems deployed in safety-critical or health-risk scenarios as AI activities subject to mandatory expert review.

In addition, the new regulations lay out systematic provisions covering the scope of the review, implementing bodies, working procedures, service support, and supervisory management.

The measures aim to move AI governance from the realm of principles into institutionalized, end-to-end ethical oversight, said Wu Qi, a senior researcher at the Pangoal Institution, noting that it is the first specialized regulation in China targeting AI ethics review.

The AI activities covered by the new measures refer to scientific research and technological development conducted within China that may pose ethical risks to human dignity, public order, life and health, the ecological environment, and other areas. Ethics reviews focus on promoting human well-being, fairness and justice, controllability and trustworthiness, transparency and explainability, accountability, and privacy.

A balance between development and security must be struck, pursuing a governance approach that combines tiered and classified oversight with tolerance for prudent innovation, according to Liang Zheng, a professor at Tsinghua University's School of Public Policy and Management and deputy director of the Institute for AI International Governance, adding that a regulatory environment that incentivizes innovation and accepts failure must be created for AI to grow into a pillar industry of tomorrow.

Companies, universities, and research institutions with strong technical capabilities should strive to develop "hard technology" for AI ethics governance, said Wei Yiming, chair of the Expert Committee on Science and Technology Ethics in the Industrial and Information Technology sector. 

There must be dedicated efforts in explainable AI, algorithmic fairness monitoring, deepfake detection, and other areas, Wei noted, urging that abstract ethical principles must be translated into quantifiable and operational technical indicators to advance the engineering and technical implementation of ethics governance.

Editor: Martin Kadiev