This is a document from the MIT AI Governance Map - using an LLM pipeline, we assessed each document in CSET Emerging Technology Observatory's AGORA dataset for coverage of risk subdomains, sectoral coverage, legislative status, affected actors and technical scope.
You can explore the full dataset and analysis on our main page or search for individual documents here.
Outlines six general guiding principles and 18 specific ethical requirements for fair, safe, and accountable AI development in China.
Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.
Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.
All other risk subdomains had no coverage in this document.
Risk Coverage Summary
The document has good coverage of approximately 10-12 subdomains, with strong focus on discrimination and fairness (1.1, 1.3), privacy protection (2.1), misinformation risks (3.1), security vulnerabilities (2.2), malicious use prevention (4.1, 4.2, 4.3), human agency (5.2), governance structures (6.5), and AI system robustness (7.3, 7.4). Coverage is concentrated in fairness, privacy, security, and governance domains.
Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.
Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.
All other sectors had no coverage in this document.
Sectors Summary
This document applies broadly across all sectors as it governs AI activities by natural persons, legal persons, and organizations engaged in AI management, research and development, supply, and use, without limiting application to specific industries. The norms are sector-agnostic ethical principles applicable to any entity working with AI technology.
Soft Law
Legislative Status Summary
This document uses predominantly voluntary and aspirational language ('should', 'encourage', 'promote') with some mandatory prohibitions ('must not', 'forbidden'). While it includes some enforcement-oriented language, it lacks specific penalties, sanctions, or enforcement mechanisms, and relies primarily on ethical guidance and self-regulation.
Proposer
Target
Enforcer
Monitor
Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems
AI Lifecycle Summary
The document explicitly covers all stages of the AI lifecycle, from planning and design through deployment and ongoing monitoring. It structures ethical norms across management, research and development, supply, and use activities, with particular emphasis on development (data quality, algorithm design) and operational monitoring (emergency protection, user feedback).
Terms relating to technical scope that are mentioned or implicitly referenced in the document
Technical Scope Summary
The document uses the general term 'AI' throughout and refers to 'AI systems,' 'AI products and services,' and 'AI technology' without distinguishing between specific types such as general purpose, task-specific, foundation models, or generative AI. There are no mentions of compute thresholds, frontier AI, or open-weight models. The scope appears to cover all AI technologies broadly.
Document data from AGORA, CSET Emerging Technology Observatory.