Resources

The dominant force in China's AI safety governance. This body holds dual identities: as a Party organ (Central Cyberspace Affairs Commission) and as a state administrative body (Cyberspace Administration of China). It is the core institution regulating generative AI, algorithmic recommendation systems, and content compliance. Its authority stems not only from technical capabilities but also from its direct connection to the Communist Party's central leadership. In the realm of generative AI, the CAC spearheaded the "Interim Measures for the Management of Generative Artificial Intelligence Services" and is responsible for implementing the algorithmic registration system. It exemplifies China’s emphasis on "controllability" and "guidance," standing at the intersection of ideological control and technical regulation.

MIIT is the main body promoting the industrialization and standardization of AI technologies in China. Its focus lies not in political sensitivity or ethics, but in integrating AI into the national industrial system. It has led the release of strategies on intelligent manufacturing, internet of vehicles, domestic chip development, and AI-empowered industries. Recently, MIIT has also joined regulatory discussions regarding model compliance, particularly in areas such as model evaluation and computing infrastructure.

NDRC serves as China's top-level macroeconomic planner. It regards AI as part of the country’s "new productive forces" and a key strategic emerging industry. It is responsible for large-scale development programs and regional innovation pilots, such as the establishment of AI innovation zones. In cross-ministry policy coordination, the NDRC often acts as a resource allocator but is less involved in the technical or safety aspects of AI regulation.

MOST is the central authority on AI ethics and scientific research governance. It established the National Governance Committee for the New Generation Artificial Intelligence and led the release of China’s AI ethical guidelines. MOST serves as a provider of "public legitimacy," offering internationally presentable frameworks for AI ethics. Additionally, it oversees the distribution of national research funds for basic AI science and major technology initiatives.

Though not directly involved in regulation, the Ministry of Education shapes the AI governance landscape through its influence on talent cultivation. In recent years, it has promoted AI curriculum in higher education, supported the founding of AI colleges at top universities, and encouraged integration between academia and industry. Given that many of China’s leading model labs (e.g., Tsinghua AIR, Peking University Zhipu) are university-based, the ministry’s policies profoundly affect the R&D ecosystem.

The "Interim Measures for the Management of Generative Artificial Intelligence Services" implemented in July 2023 exemplify China’s multi-agency approach to AI regulation. The policy was jointly issued by seven central government bodies:

• The Cyberspace Administration of China (CAC) led the drafting to ensure political alignment and content safety.

• The National Development and Reform Commission (NDRC) aligned the policy with national digital infrastructure and innovation goals.

• The Ministry of Industry and Information Technology (MIIT) contributed technical norms concerning training data, model development, and computing resources.

• The Ministry of Science and Technology (MOST) incorporated ethical governance frameworks.

• The Ministry of Education focused on AI talent and educational integration.

• The Ministry of Public Security addressed security risks such as deepfakes and fraud.

• The National Radio and Television Administration ensured generated content aligned with national media and cultural standards.

AI-specific legislation in China falls under the authority of the National People’s Congress (NPC) and its Standing Committee’s Legislative Affairs Commission. Though no AI Act exists yet, the NPC has passed three foundational laws for digital governance:

• Cybersecurity Law (2017): Defines basic boundaries for data flows and infrastructure protection;

• Data Security Law (2021): Establishes classification and graded protection of data;

• Personal Information Protection Law (2021): Addresses automated decision-making, data minimization, and user rights.

Together, these form a legal shield for AI governance, especially around platform liability and user rights.

A strategic alliance of leading Chinese AI research institutions. It is not a formal government agency, lacks a board or administrative hierarchy, but includes many of the most influential government-affiliated think tanks and laboratories. It functions as China’s de facto hub for AI safety discourse and international engagement.