AI Governance in China
Five enacted regulations. Three data laws. One filing system you need to understand.
China’s Three-Layer AI Regulatory Stack
Click each layer to see the regulations, laws, and standards within it.
The AI Law is on the NPC legislative agenda. No draft has been published publicly.
Four Eras of Chinese AI Governance
From open investment to regulatory crossroads in under a decade.
Framework adapted from Carnegie Endowment for International Peace
Minimal regulation, massive state investment. China captured 48% of global AI equity funding in 2017. The “New Generation AI Development Plan” set a goal to lead the world in AI by 2030.
The Cyberspace Administration of China (CAC, 网信办) emerged as the dominant AI regulator. Algorithm Recommendation Provisions and Deep Synthesis Provisions enacted. PIPL and DSL passed within months of each other.
ChatGPT shocked Chinese regulators. The draft GenAI Measures were strict, but the final version softened after industry pushback. Filing requirements replaced pre-launch approval for services using already-filed models.
DeepSeek’s breakthrough shifted the narrative. Content labeling rules took effect. Ethics review measures formalized three-tier review committees. The “AI+” industrial initiative signals growth priorities alongside regulation. A unified AI Law (人工智能法) is on the NPC legislative agenda, though no draft has been published publicly as of May 2026.
What Do You Need to Know?
Data Triad: PIPL, DSL, CSL
How China’s three data laws apply to AI systems
See the three data laws →CAC Filing System
Which of the three filing types does your AI service need?
Identify your filing type →Generative AI Measures
The cornerstone August 2023 GenAI regulation
Read the GenAI Measures →Content Rules and Labeling
Mandatory labeling, watermarking, and metadata requirements
Check labeling specs →TC260 National Standards
83 published standards and technical safety thresholds
Browse 83 standards →AI Ethics Review
Mandatory ethics committees and three-tier review system (2026)
Review the three-tier system →Cross-Border Data Transfer
Three primary mechanisms for cross-border AI data transfer
Map your transfer mechanism →China vs. Global Frameworks
Side-by-side with EU, US, India, Japan, and Singapore
Compare 6 jurisdictions →Sector Regulators
PBoC, CSRC, and SAMR sector-specific AI rules
Find your sector rules →Templates and Tools
Compliance checklists and decision trees
Download tools →Key Agencies and Standards Bodies
Tap a card to see specific responsibilities.
Lead AI regulator. Manages all three filing registries and content enforcement.
Telecommunications, AI chip controls, and industrial AI standards.
AI unfair competition, pricing algorithms, and consumer protection.
Standards committee under SAC (not a regulatory body). Develops AI safety and security technical standards that regulators reference in enforcement.
Science and technology ethics review, AI research governance.
Cybersecurity law enforcement, network security inspections.
Recent Enforcement Campaigns
These are not hypothetical risks. Enforcement is active and public.
Maximum Penalty Ranges
Revenue-based penalties apply to each law independently.
Which Filing Type Do You Need?
Three questions to identify your most likely filing obligation.
Global AI Governance at a Glance
Six jurisdictions. Four dimensions. Scroll horizontally on mobile.
| Dimension | China | EU | US | India | Japan | Singapore |
|---|---|---|---|---|---|---|
| Risk Classification | Activity-based | 4-tier risk | Voluntary | 6 context-specific categories | Use-case-based | Voluntary, risk-proportionate |
| Pre-Launch Filing | Mandatory (CAC) | Conformity assessment | None | None (voluntary) | None (soft law) | None (AI Verify voluntary) |
| Content Controls | Mandatory keyword lists | Transparency labels | None | Mandatory labeling, 2-3hr takedowns | Voluntary watermarking | Voluntary provenance |
| Penalties | Up to 10% revenue | Up to 7% turnover | Sector-specific | Up to INR 250 crore (DPDPA) | No AI-specific penalties | No AI-specific penalties |
Templates and Decision Trees
Download tools built for China’s specific regulatory requirements.
CAC Filing Readiness Assessment
12-question self-assessment to determine your filing obligations.
China AI Compliance Checklist
Cross-law checklist covering PIPL, DSL, CSL, and all five AI regulations.
Cross-Border Transfer Decision Tree
Map your data flows to the correct transfer mechanism.
Need Help with China AI Compliance?
TJS advisors help multinational teams with CAC filing preparation, cross-border data transfer structuring, and multi-jurisdiction compliance mapping.
Talk to a TJS Advisor →