New-Generation AI Development Plan
📑 Legal hierarchy: Level 4 · Normative document (a development plan issued by the State Council) | Issuance: State Council | Document number: Guofa [2017] No. 35 | Issued: 2017-07-20 | Character: soft law · strategic plan
⚠️ Hierarchy note: This document is a development plan issued by the State Council (a normative document, Level 4), not an administrative regulation (Level 2). The Plan sets strategic objectives and policy directions; it does not itself impose duties on firms. But as the source of all subsequent AI departmental rules, subnational AI legislation, and industrial policies, it is indispensable for understanding Chinese AI governance. See Index of Chinese Rules.
Chinese Summary
Section titled “Chinese Summary”The New-Generation AI Development Plan (Guofa [2017] No. 35) was signed by the State Council on 2017-07-08 and issued on 2017-07-20. It is the top-level design of China’s AI strategy and the first national AI plan.
Significance: every subsequent Chinese rule, industrial policy, or subnational legislation on AI traces back to this Plan. Principles such as “inclusive and prudent,” “classified and graded,” and “ethics norms” are officially articulated here for the first time.
”Three-Step” Strategic Goals
Section titled “”Three-Step” Strategic Goals”| Milestone | Core indicator | Industry scale |
|---|---|---|
| 2020 | AI technology and application at the same level as the world frontier | Core industry over RMB 150 billion; related industries over RMB 1 trillion |
| 2025 | Some technology and applications reach world-leading levels | Core industry over RMB 400 billion; related industries over RMB 5 trillion |
| 2030 | Theory, technology, and applications overall at world-leading levels; becoming a major global AI innovation center | Core industry over RMB 1 trillion; related industries over RMB 10 trillion |
Main Tasks
Section titled “Main Tasks”1. Build an open, collaborative AI S&T innovation system
Section titled “1. Build an open, collaborative AI S&T innovation system”- A four-in-one combination of basic theory + key common technologies + innovation platforms + talent pool.
- National New-Generation AI Open Innovation Platforms (the 15+ platforms subsequently hosted by Baidu, Alibaba, Tencent, SenseTime, iFlytek, Huawei, etc.).
2. Cultivate a high-end, efficient intelligent economy
Section titled “2. Cultivate a high-end, efficient intelligent economy”- Drive integration of AI with manufacturing, agriculture, finance, logistics, and other industries.
- Cultivate leading AI firms and industry clusters.
3. Build a safe, convenient intelligent society
Section titled “3. Build a safe, convenient intelligent society”- Deep application in education, healthcare, elder care, urban management and more.
- Smart cities, intelligent transportation, intelligent public security.
4. Strengthen civil-military integration
Section titled “4. Strengthen civil-military integration”Promote two-way transfer of AI technology between civilian and military uses.
5. Build a ubiquitous, secure, and efficient intelligent infrastructure system
Section titled “5. Build a ubiquitous, secure, and efficient intelligent infrastructure system”- 5G / big data / high-performance computing / IoT.
6. Forward-deploy major S&T projects in new-generation AI
Section titled “6. Forward-deploy major S&T projects in new-generation AI”Safeguards (Including Governance Foreshadowing)
Section titled “Safeguards (Including Governance Foreshadowing)”Laws, regulations, and ethics norms
Section titled “Laws, regulations, and ethics norms”制定人工智能相关法律、伦理和社会问题框架，研究自动驾驶、服务机器人等应用领域的安全管理法规。
— English gloss — Formulate a framework for legal, ethical, and social issues related to AI; study safety administration regulations for application areas such as autonomous driving and service robots.
→ This foreshadowing was gradually operationalized by:
- 2019 New-Generation AI Governance Principles (eight items);
- 2022–2026 the series of departmental rules on algorithms / deep synthesis / generative AI / labeling / anthropomorphic interaction;
- 2024–2025 AI Safety Governance Framework 1.0 / 2.0.
Technical-standards system
Section titled “Technical-standards system”逐步建立并完善 AI 基础共性、互联互通、行业应用、网络安全、隐私保护等技术标准。
— English gloss — Progressively establish and refine technical standards for AI in foundational commonality, interoperability, industry application, cybersecurity, and privacy protection.
→ Operationalized by the TC260 series of standards (including TC260-003-2024).
Safety supervision
Section titled “Safety supervision”Establish AI safety monitoring and early-warning mechanisms and a dynamic R&D-and-application assessment-and-evaluation mechanism.
Human-resource development
Section titled “Human-resource development”- Lifelong-learning system;
- Vocational retraining.
Structural Impact on Subsequent Regulation
Section titled “Structural Impact on Subsequent Regulation”| Plan’s policy phrasing | Subsequent rule that operationalizes it |
|---|---|
| ”Classified and graded supervision” | TC260-003-2024 31 risk categories + AI Safety Governance Framework |
| ”Agile governance” | 2019 Principles made this explicit; subsequent “Interim Measures” naming tradition |
| ”Safety administration regulations” | Deep Synthesis Provisions / Generative AI Interim Measures / Labeling Measures and others |
| ”National innovation platforms” | 15+ National New-Generation AI Open Innovation Platforms hosted by leading firms |
| ”Civil-military integration” | Subsequent participation by the CMC S&T Commission as a co-signatory of the Science and Technology Ethics Review Measures |
Comparison with Contemporaneous International Policy
Section titled “Comparison with Contemporaneous International Policy”- US at the same time: Obama 2016 “Preparing for the Future of AI” and “National AI R&D Strategic Plan.”
- EU at the same time: no dedicated AI strategy yet; Artificial Intelligence for Europe only in 2018.
→ China’s strategic AI document preceded the EU’s by one year.
Source Text and Archival Copies
Section titled “Source Text and Archival Copies”| Source | Link |
|---|---|
| Chinese (State Council) | gov.cn/…/2017/07/20/content_5211996.htm |
| Chinese (archived copy) | archives/china/new-gen-ai-development-plan-2017-07-20.html |
| English | multiple unofficial translations (CSET, DigiChina, etc.) |
Version History
Section titled “Version History”| Date | Event |
|---|---|
| 2017-07-08 | Deliberated by State Council executive meeting |
| 2017-07-20 | Officially issued |
| 2020 | Stage 1 milestone |
| 2025 | Stage 2 milestone |
| 2030 | Stage 3 milestone |