China’s Ministry of Industry and Information Technology, along with nine other governmental departments, has introduced new trial rules governing the ethics of artificial intelligence (AI) in scientific and technological activities. Titled the ‘Measures for AI Science and Technology Ethics Review and Services (Trial)’, these regulations aim to establish a comprehensive framework for ethics governance in AI, focusing on promoting responsible innovation that is fair, just, and safe.
The newly issued measures cover a wide range of AI-related activities conducted within China that may present ethical risks, including research and technology development. These risks are particularly pertinent to issues of human dignity, public order, health, the ecological environment, and sustainable development. The document emphasizes that ethical considerations must permeate all stages of AI activities, setting forth principles such as promoting human well-being, respecting life and rights, ensuring fairness, exercising reasonable risk control, and safeguarding privacy and security.
Support measures outlined in the guidelines call for a robust AI ethics standards system. This includes the development of international, national, and industry standards, alongside enhanced risk monitoring and assessment protocols. The measures also seek to support small and micro-enterprises, advance research on ethics review processes, and improve public education and talent training related to AI ethics. The establishment of AI ethics committees within universities, research institutions, and businesses is mandated to ensure compliance with the new standards.
Local authorities and relevant departments are encouraged to create specialized ethics review and service centers, which can offer training, consulting services, and conduct reviews. However, these centers are prohibited from performing both review and re-examination of the same AI activity to maintain impartiality in the assessment process.
The application and review protocols are laid out clearly, encompassing general, simplified, expert re-examination, and emergency procedures. Reviews are to prioritize human well-being, fairness, controllability, transparency, traceability of responsibility, and privacy protection. The decision on a review must be issued within 30 days after submission, with a provision for extension in complex cases. Emergency reviews are expected to be completed within 72 hours.
Furthermore, the guidelines specify expert re-examination for certain listed activities that significantly influence human behavior, emotions, or health, as well as those that involve highly autonomous decision-making systems with safety or health implications. The list of activities subject to these stringent reviews will be dynamically adjusted as necessary, reflecting the evolving landscape of AI technologies.
The document also stipulates that violations of these measures may be subject to legal scrutiny under various laws, including the Cybersecurity Law, the Data Security Law, the Personal Information Protection Law, and the Science and Technology Progress Law. The measures are effective immediately upon issuance, signaling a notable shift in China’s approach to regulating AI ethics.
This move comes as global concerns grow regarding the ethical implications of AI technology. By instituting these rules, China aims to foster a culture of ethical awareness in AI development, potentially influencing how similar regulations evolve worldwide. As AI continues to permeate various sectors, the establishment of clear ethical guidelines may play a crucial role in shaping its future trajectory.
See also
OpenAI’s Rogue AI Safeguards: Decoding the 2025 Safety Revolution
US AI Developments in 2025 Set Stage for 2026 Compliance Challenges and Strategies
Trump Drafts Executive Order to Block State AI Regulations, Centralizing Authority Under Federal Control
California Court Rules AI Misuse Heightens Lawyer’s Responsibilities in Noland Case
Policymakers Urged to Establish Comprehensive Regulations for AI in Mental Health
