Global AI Governance Initiative
📑 Legal hierarchy: Level 4 · Normative document · externally facing position document | Issuance: People’s Republic of China (proposed by President Xi Jinping) | Released: 2023-10-18 | Character: soft law · international political declaration
⚠️ Hierarchy note: This document is an externally facing policy-position document, not a domestic law or departmental rule. It does not directly impose duties on domestic firms, but it is China’s de facto AI-governance position document in the UN, G20, BRICS, and China–US / China–EU bilateral dialogues, and informs AI-related diplomacy and international cooperation. See Index of Chinese Rules.
Chinese Summary
Section titled “Chinese Summary”The Global AI Governance Initiative was proposed by President Xi Jinping on 2023-10-18 in the keynote address at the opening of the Third Belt and Road Forum for International Cooperation, and subsequently released jointly by the Ministry of Foreign Affairs and the Cyberspace Administration of China.
Positioning: the official externally facing articulation of China’s AI-governance stance, setting out the Chinese approach systematically along three axes (development / safety / governance).
Core Propositions
Section titled “Core Propositions”Organized around AI development, safety, and governance:
On AI Development
Section titled “On AI Development”- Uphold people-centered, AI for good (以人为本、智能向善) — guide AI toward directions favorable to the progress of human civilization.
- Mutual respect, equality, and mutual benefit — oppose drawing ideological lines or forming exclusive blocs; oppose malicious obstruction of other countries’ AI development.
- Open cooperation — support AI development in countries of the Global South; narrow the digital divide.
On AI Safety
Section titled “On AI Safety”- Promote a risk-level testing-and-assessment system;
- Implement agile governance;
- Classified and graded management;
- Rapid and effective response;
- R&D actors to continuously enhance AI explainability and predictability;
- Improve the authenticity and accuracy of data;
- Ensure AI remains under human control.
On AI Governance
Section titled “On AI Governance”- Under the UN framework: actively support discussion of establishing an international AI governance body under the UN framework.
- Representation of developing countries: enhance the representation and voice of developing countries in AI global governance.
- Cooperation and aid for developing countries.
- Opposition to technological monopolies and export controls (with explicit and implicit reference to US chip controls).
Mapping onto Domestic Rules
Section titled “Mapping onto Domestic Rules”The Initiative’s domestic counterparts are China’s existing governance rules:
| Initiative proposition | Corresponding domestic rule |
|---|---|
| ”People-centered, AI for good” | New-Generation AI Governance Principles + Anthropomorphic Interaction Services Measures |
| ”Agile governance” + “classified and graded” | Scenario-specific departmental rules + AI Safety Governance Framework |
| ”Explainability, predictability” | TC260-003-2024 |
| ”Authenticity and accuracy of data” | Generative AI Interim Measures Article 7 |
| ”Under human control” | As above + Deep Synthesis Provisions |
International Follow-up
Section titled “International Follow-up”- 2024-07: The UN General Assembly adopted China’s proposed resolution on “Enhancing International Cooperation on AI Capacity Building.”
- 2024-05: The Shanghai World AI Conference (WAIC) released the Shanghai Declaration.
- 2025-07: China, together with Global South partners, proposed an “AI capacity-building action plan.”
- 2025-09: Vice Foreign Minister Ma Zhaoxu spoke at the UN high-level meeting on global AI governance dialogue.
- Reiterated in G20, BRICS+, APEC, and the Forum on China-Africa Cooperation venues.
Comparison with International Initiatives
Section titled “Comparison with International Initiatives”| Document | Initiating party | Released | Emphasis |
|---|---|---|---|
| Global AI Governance Initiative | China | 2023-10 | Balance development and safety + developing-country representation + opposition to technology monopolies |
| Bletchley Declaration | UK / multi-country | 2023-11 | Frontier AI risks, 28 signatories (including China) |
| Seoul Declaration | Korea / multi-country | 2024-05 | International AISI network |
| G7 Hiroshima Process | G7 | 2023 | Code of conduct |
| Paris AI Summit Declaration | France | 2025-02 | 60+ signatories; US and UK declined |
The Chinese Initiative contrasts with Bletchley / Seoul: China emphasizes developing-country representation and opposition to export controls; the UK / US / EU side emphasizes frontier-risk assessment and red lines.
Source Text and Archival Copies
Section titled “Source Text and Archival Copies”| Source | Link |
|---|---|
| Chinese (CAC) | cac.gov.cn/2023-10/18/c_1699291032884978.htm |
| Chinese (MFA) | mfa.gov.cn/…/t20231020_11164831.shtml |
| Chinese (archived copy) | archives/china/global-ai-governance-initiative-2023-10-18.html |
| English (official MFA translation) | — |
| Press Q&A | mfa.gov.cn/fyrbt/…/t20231018_11162801.shtml |
Version History
Section titled “Version History”| Date | Event |
|---|---|
| 2023-10-18 | Announced at the Third Belt and Road Forum |
| 2024-07 | UN resolution on AI capacity-building cooperation (in the spirit of the Initiative) |
| 2025-09 | Ma Zhaoxu reiterates the position at the UN global AI governance dialogue |