China’s Ministry of Industry and Information Technology (MIIT), in coordination with nine other government agencies, has issued the “Administrative Measures for the Ethical Review and Services of Artificial Intelligence Science and Technology (Trial).” The new regulations mark a significant shift in China’s AI governance, moving from a focus primarily on content security to a comprehensive, institutionalized ethical compliance system.
The framework establishes a three-tier structure for ethical oversight: internal ethics committees within organizations, external service centers, and government-led expert reviews. All universities, research institutions, and companies engaged in AI development are now required to establish internal ethics committees. For high-risk projects—defined as those affecting public opinion, psychological behavior, or involving highly automated decision-making—mandatory entry into a government-led expert review process is required before the project can proceed.
The measures introduce a dual-track access model combining algorithm filing with ethical evaluation. Before commencing a project, applicants must submit detailed materials including technical plans, data sources, algorithmic mechanisms, and ethical risk assessments. Review bodies must issue decisions within 30 days. Crucially, the regulations mandate ongoing monitoring, meaning projects can be suspended or terminated if risk conditions change. The rules also address labor protections, requiring algorithm auditing for platform-based sectors like ride-hailing to incorporate human override functions, preventing the “algorithmic exploitation” of workers.
This comprehensive regulatory framework underscores Beijing’s determination to maintain tight control over the rapid development and deployment of artificial intelligence. Unlike the more fragmented and industry-led approaches seen in Western jurisdictions, China is establishing a centralized, state-directed system for ethical oversight. This approach reflects a broader governance philosophy that prioritizes social stability and alignment with national strategic goals over unfettered technological innovation.
The requirement for all AI developers to establish internal ethics committees represents a significant compliance burden, particularly for smaller startups and research institutions. This mandate will likely drive the creation of a new industry focused on AI ethics consulting and compliance services, as organizations scramble to meet the stringent new requirements. The mandatory government-led expert review for high-risk projects further centralizes authority, giving the state direct veto power over the most sensitive and potentially disruptive AI applications.
The inclusion of labor protections within the ethical framework is a notable development, reflecting growing concerns about the impact of algorithmic management on gig economy workers. By requiring human override functions and auditing for “algorithmic exploitation,” regulators are attempting to mitigate the negative social consequences of AI deployment in sectors like ride-hailing and food delivery. This focus on the socioeconomic impacts of AI distinguishes China’s approach from frameworks that focus more narrowly on technical safety and data privacy.
For international observers, these new regulations provide valuable insight into the evolving landscape of AI governance in China. While the stringent oversight mechanisms may slow the pace of certain types of innovation, they also provide a clear and predictable regulatory environment for developers. This clarity could ultimately prove advantageous, allowing Chinese companies to deploy AI solutions with greater confidence, knowing they have met the state’s ethical requirements.
However, the broad definitions of “high-risk projects” and the subjective nature of ethical evaluations introduce a degree of uncertainty. The implementation and enforcement of these rules will be closely watched, as they will determine the practical impact on China’s vibrant AI ecosystem. If the review process becomes overly bureaucratic or politicized, it could stifle the very innovation Beijing seeks to foster. Conversely, a well-calibrated approach could establish China as a global leader in responsible AI development.
Ultimately, the “Administrative Measures for the Ethical Review and Services of Artificial Intelligence Science and Technology” represent a critical milestone in the maturation of China’s AI industry. By institutionalizing ethical oversight and integrating it directly into the development lifecycle, Beijing is asserting its authority over the trajectory of this transformative technology, ensuring it serves the broader interests of the state and society.