China - AI Development Plan (2017)

Overview

The New Generation Artificial Intelligence Development Plan was completed on July 8, 2017, and publicly released on July 20, 2017, under State Council Notice No. 35 (2017). It functions as China's top-level national strategy for artificial intelligence development rather than a statute, establishing a three-step roadmap targeting 2020, 2025, and 2030. The authoritative text is available on the State Council Policy Repository and the official State Council Release Page.

Definitions

Within the Plan, "new generation artificial intelligence" refers to technologies and systems with capabilities such as deep learning, cross-media perception and reasoning, hybrid human-machine intelligence, swarm intelligence and autonomous decision-making. The document uses policy terms rather than technical legal definitions; it frames AI activities into categories such as foundational research (basic theory and core algorithms), common key technologies (chips, sensors, operating platforms), application systems (industry-specific deployments), intelligent infrastructure (communications, data and high-performance computing), and ecosystem functions (standards, testing, talent and finance). Additionally, the Plan introduces operational constructs: the "1" major national science and technology AI project, and the "N" set of supporting projects and programs across ministries and provincial authorities.

Governance and Institutional Framework

The Plan establishes an institutional architecture for coordination and implementation. It tasks the Ministry of Science and Technology with hosting an AI planning promotion office and creating a strategic advisory committee to provide consultative guidance and evaluation. Other central bodies with assigned roles include the Cyberspace Administration (data, security, content and cross-border data concerns), the Ministry of Industry and Information Technology (standards and industrial policy), the National Development and Reform Commission (planning and major projects), the Ministry of Finance (funding guidance), and provincial governments for local implementation. The Plan emphasizes a government-market division of labor: government to plan, coordinate, convene resources and set security/ethical boundaries; markets and enterprises to lead product innovation and commercialization. For central coordination and public dissemination the Plan is hosted on the State Council portal; implementation instructions and follow-up measures are to be issued by relevant departments. See the Plan and implementation office reference at State Council (Chinese) and the English translation hosted by research institutions at New America — translation (PDF).

Key Focus Areas

The Plan organizes six major task clusters. (1) Scientific and technological innovation system: strengthen basic theory, key common technologies, innovation platforms, and high-end talent pools, with open-source sharing encouraged. (2) Intelligent economy and industry: cultivate AI core industries, spur intelligent transformation of traditional sectors (manufacturing, agriculture, transport, finance, healthcare), and establish industrial clusters. (3) Intelligent society: accelerate AI adoption in public services (healthcare, education, elderly care, environment, urban operations, judicial assistance), while highlighting safety, privacy and social governance needs. (4) Military-civil integration: promote two-way conversion of military and civilian AI applications and shared innovation resources. (5) Infrastructure: build pervasive networks (5G and beyond), national data sharing platforms, and high-performance computing ecosystems. (6) Major projects and foresight: deploy a "1+N" program of national-level science and technology projects to address foundational bottlenecks (brain-like computing, autonomous systems, group intelligence, knowledge computing engines). Across these areas the Plan also prioritizes standards, intellectual property protection, pilot demonstrations, and ethical/legal research addressing traceability and accountability.

Implementation Framework

Implementation is to be realized through coordinated central-ministerial tasks, provincial/local rollout, public–private partnerships, targeted funding instruments and dedicated major national projects. The Plan instructs formation of annual and phased implementation plans, establishment of pilot demonstration zones, integration of local industrial resources and investment pools, and scientific project clusters tied to national R&D mechanisms (e.g., National Key R&D Program, NSFC). It stresses fiscal and financial guidance (funds, venture capital incentives), standards development led by MIIT and standards bodies, and development of testing and safety evaluation institutions to provide conformity assessment for critical AI applications. The Plan asks ministries to allocate responsibilities, produce progress metrics, and update the Plan dynamically in response to technological advances.

Monitoring and Evaluation

The Plan mandates periodic monitoring and evaluation mechanisms, including annual assessments and a mid-term review process to check progress against the 2020/2025/2030 milestones. The AI planning office under the Ministry of Science and Technology is assigned coordination responsibility for aggregating national performance indicators (industry scale, technology breakthroughs, talent training, standards adoption), while specific ministries measure sectoral rollout and pilot outcomes. The Plan calls for the development of safety assessment and risk analysis capabilities, creation of metrics for social impacts (employment transformation, privacy and security incidents), and an evidence base to inform dynamic updating of priorities and funding. It also signals the need for public reporting, demonstration case studies and dissemination of replicable policy practices across provinces.

Penalties, Liability, and Appeals

As a high-level strategy document, the Plan itself does not create a separate enforcement regime or criminal penalties. Instead, it instructs research and formulation of laws, regulations and ethical norms (including traceability, accountability and responsibility assignment) and signals that enforcement will rely on existing legal instruments and subsequent ministerial rules. In practice this means that administrative penalties, license restrictions, funding withdrawal, market access limitations and criminal liability (under pre-existing statutes) can be applied to AI actors who violate specific safety, cybersecurity, data protection or other substantive legal obligations once such statutory rules are adopted. The Plan recommends building traceability and accountability systems to underpin civil liability and administrative measures, and it expects ministries to specify appeals and review mechanisms in implementing rules and pilot programs.

Relationship to Other Instruments

The Plan sits at the apex of China’s AI policy architecture and is explicitly linked to other national initiatives such as industrial policies, the National Key R&D Program, "Made in China 2025" (industrial upgrading), "Internet+" initiatives and subsequent AI-related departmental rules and local government plans. It provides strategic direction rather than operational regulation, and subsequent sectoral regulatory acts (data security, personal information protection, cybersecurity rules, sectoral safety standards) and local implementation plans translate the Plan into concrete obligations. The Plan also set expectations for standard-setting bodies and industry consortia to develop technical standards, conformity assessment schemes and testing capabilities in coordination with MIIT and other departments.

International Alignment

The Plan emphasizes international cooperation on research, standards and governance: it advocates active participation in global R&D networks, norms development, international standard-setting organizations and collaborative projects. While the Plan emphasizes China’s strategic autonomy and national security, it explicitly calls for open-source sharing and global R&D integration to mobilize worldwide innovative resources. That dual posture has guided later Chinese engagement in ISO/ITU and other international standards fora and multilateral policy dialogues on AI ethics and governance.

Implementation Timeline

Sources and References

Official Policy Document: Notice of the State Council on Issuing the Development Plan for the New Generation of Artificial Intelligence (国发〔2017〕35号).

Official Publication Page: State Council Full Text Release of New Generation Artificial Intelligence Development Plan.