GIRAI logo
China flag

Global Index Score

43.58

Global: 42ndRegional: 7thIncome Group: 6th

China's index score lands +10.6 against the Asia and Oceania average (33.0) and +11.7 against Upper middle income peers (31.9). It scores ahead of 69% of the 135 economies assessed this edition. Published score 43.6 reflects a URAI adjustment from 48.4 (4 government-misuse evidence items on file). The profile rests on 57 documented evidence items spanning policy frameworks, government initiatives and civil society activity.

How China Performs Across the Five Dimensions

GIRAI evaluates countries across five core dimensions that capture the social, ethical, and institutional impacts of artificial intelligence.

AI Policy

50.90

Avrg. Score

25.65

Global: 26thRegional: 2nd

Enabling Conditions

48.51

Avrg. Score

55.27

Global: 95thRegional: 21st

Civil Society Engagement

33.27

Avrg. Score

17.68

Global: 26thRegional: 6th
1

Inclusion and Diversity

43.97

Index Score

33.21

Avrg. Dimension score

Global: 44thRegional: 8th

China runs +10.8 against the regional average (33.2) on Inclusion and Diversity. Strongest indicators: Gender Equality (56.1), Children's Rights (47.0). Weakest: Cultural and Linguistic Diversity (18.2). This dimension ranks 44th globally — behind the country's overall position (42nd). 10 evidence items on file, including "AI Safety Governance Framework 2.0 (China)" and 2 other documented items.

2

Ethics and Sustainability

53.43

Index Score

31.00

Avrg. Dimension score

Global: 28thRegional: 3rd

China runs +22.4 against the regional average (31.0) on Ethics and Sustainability. Strongest indicators: Environmental Impact (74.6), Transparency and Explainability (68.9). Weakest: Human Oversight and Determination (43.2), Fairness and Non-discrimination (61.4). This dimension ranks 28th globally — ahead of the country's overall position (42nd). 17 evidence items on file, including "AI Safety Governance Framework 2.0 (China)" and 1 other documented items.

3

Labour and Skills

49.23

Index Score

36.03

Avrg. Dimension score

Global: 43rdRegional: 9th

China runs +13.2 against the regional average (36.0) on Labour and Skills. Strongest indicators: AI Literacy (78.8), Reskilling/Upskilling Initiatives (68.9). Weakest: Labour Protections (38.6). This dimension ranks 43rd globally — behind the country's overall position (42nd). 7 evidence items on file, including "AI curriculum for basic education".

4

Trust and Safety

64.16

Index Score

38.45

Avrg. Dimension score

Global: 26thRegional: 2nd

China runs +25.7 against the regional average (38.4) on Trust and Safety. Strongest indicators: Access to Redress and Remedy (86.4), Safety and Security (68.2). Weakest: AI-facilitated Misinformation and Violence (48.9), Impact Assessments (56.8). This dimension ranks 26th globally — ahead of the country's overall position (42nd). 13 evidence items on file, including "AI Safety Governance Framework 2.0 (China)" and 2 other documented items.

5

AI Use in Public Service

31.29

Index Score

30.02

Avrg. Dimension score

Global: 56thRegional: 14th

China runs +1.3 against the regional average (30.0) on AI Use in Public Service. Strongest indicators: Public Sector Skills Development (65.5), Public Disclosure of Government Algorithmic Systems (9.1). Weakest: Public Procurement (0.0). This dimension ranks 56th globally — behind the country's overall position (42nd). 10 evidence items on file, including "Internet Information Service Algorithm Filing System" and 1 other documented items.

Comparing Results Across Editions

Explore how country performance, scores, and governance indicators have changed between GIRAI editions.

14 indicators

Indicator2024 Edition2026 Edition
Gender EqualityNo FrameworkNon-Binding Framework
Children's RightsDraftNon-Binding Framework
Cultural and Linguistic DiversityNo FrameworkNo Framework
Fairness and Non-discriminationNon-Binding FrameworkBinding Framework
Transparency and ExplainabilityNon-Binding FrameworkBinding Framework
Human Oversight and DeterminationNon-Binding FrameworkNon-Binding Framework
Labour ProtectionsBinding FrameworkBinding Framework
Reskilling/Upskilling InitiativesNo FrameworkNon-Binding Framework
Safety and SecurityNo FrameworkNon-Binding Framework
Access to Redress and RemedyDraftBinding Framework
Impact AssessmentsNon-Binding FrameworkNon-Binding Framework
Public Sector Skills DevelopmentNo FrameworkNon-Binding Framework
Public ProcurementNo FrameworkNo Framework
Government Mechanisms for CSO Inclusion in AI Policy and GovernanceNon-Binding FrameworkNon-Binding Framework

Indicator definitions changed between editions; this table compares evidence presence and status, not directly comparable scores. A — indicates the item is not applicable or was not assessed for that edition.

Compare responsible AI performance

Explore how countries and regions perform across GIRAI's governance dimensions, scores, and structural indicators.

Compare
🇨🇳China

43.58

Global 42ndRegional 7th
Asia and Oceania

32.98

Regional average

Dimension scores

Inclusion and Diversity
China
44.0
Asia and Oceania
33.2
Ethics and Sustainability
China
53.4
Asia and Oceania
31.0
Labour and Skills
China
49.2
Asia and Oceania
36.0
Trust and Safety
China
64.2
Asia and Oceania
38.4
AI Use in Public Service
China
31.3
Asia and Oceania
30.0

Indicators

37 of 37
IndicatorChinaAsia and Oceania
Inclusion and Diversity7
APGender Equality
56.124.5
APChildren's Rights
47.027.8
APCultural and Linguistic Diversity
18.225.9
CSOCivil Society Engagement in Inclusion and Diversity
23.115.4
ECEgalitarian Democracy
7.533.2
ECSocioeconomic Inclusion in Connectivity
74.849.2
ECGender Inclusion in Connectivity
91.978.0
Ethics and Sustainability6
APFairness and Non-discrimination
61.426.1
APTransparency and Explainability
68.933.5
APHuman Oversight and Determination
43.226.2
APEnvironmental Impact
74.618.0
CSOCivil Society Engagement in Ethics and Sustainability
46.120.8
ECEnvironmental Performance
38.240.9
Labour and Skills6
APLabour Protections
38.67.7
APReskilling/Upskilling Initiatives
68.941.4
APAI Literacy
78.839.0
CSOCivil Society Engagement in Labour and Skills
0.018.6
ECLabour Rights
0.039.0
ECPopulation Digital Readiness
79.771.3
Trust and Safety11
APSafety and Security
68.233.8
APAccess to Redress and Remedy
86.423.2
APImpact Assessments
56.827.9
APAI-facilitated Misinformation and Violence
48.925.5
CSOCivil Society Engagement in Trust and Safety
57.113.6
ECData Protection and Privacy
75.056.4
ECData Sharing and Access
37.568.3
ECConsumer Protection
58.072.3
ECCybersecurity
91.376.8
ECRule of Law
39.148.2
ECGlobal Peace
61.768.4
AI Use in Public Service7
APPublic Sector Skills Development
65.532.0
APPublic Disclosure of Government Algorithmic Systems
9.111.1
APPublic Procurement
0.014.8
CSOGovernment Mechanisms for CSO Inclusion in AI Policy and Governance
40.020.0
ECCivil Society Oversight
6.747.9
ECPublic Service Delivery
77.576.9
ECRight to Information
39.539.6
Score adjustment
URAIUnacceptable Risk AI Systems
×0.904 government-misuse items×0.98Regional average
APAI Policy
CSOCSO Engagement
ECEnabling Conditions
Misuse Of AI

Unacceptable Risk AI Systems

Evidence related to government use of AI systems associated with unacceptable-risk categories.

Contact GIRAI
  • Mass Biometric Surveillance

    ‘Sharp Eyes’ Surveillance Program

    The Sharp Eyes and SkyNet programs qualify as real-time remote biometric identification for law enforcement purposes in publicly accessible spaces. Sharp Eyes (launched in 2015) with the goal of achieving 100% coverage of public spaces by 2020 integrating automated facial recognition algorithms into networks of 200+ million cameras. SkyNet (since 2005) which focuses on installing cameras connected to facial recognition algorithms uses facial recognition data from building complexes shared with local police to build population databases. Both projects are ongoing and fully operational utilizing facial recognition algorithms to capture analyze and process visual and behavioral data in real-time.

    View Trusted Source
  • Criminal Justice & Law Enforcement

    AI-Powered Anti-Corruption Drive

    This case qualifies as an unacceptable risk because it automates legal investigation undermining due process and privacy. According to the report the system uses algorithmic profiling to map social networks and hidden equity nesting flagging individuals based on patterns (section 'AI and the Future of Anti-Corruption in China'). It fits individual risk assessment by using dynamic risk modelling to predict graft creating a digital panopticon where officials are judged by a black-box system.

    View Trusted Source
  • Social Scoring

    AI deployment in the credit reporting by the People's Bank of China

    The report confirms that the central government via the People's Bank of China (PBOC) actively promotes the use of AI in credit scoring. The PBOC states its intent to 'deepen the application of big data artificial intelligence blockchain and other new technologies in the credit reporting field' to fully unleash data value and achieve high-quality development and security. By licensing and regulating private agencies like Baihang and Pudao to profile individual behaviors using 'big data' algorithms the state creates a standardized system where social and financial conduct determines an individual's access to critical services and life opportunities.

    View Trusted Source
  • Cyberattacks & Influence Operations

    AI-orchestrated cyber espionage campaign

    Anthropic reported disrupting an espionage campaign that used their AI tool. Anthropic claims that the actor was 'a Chinese state-sponsored group' with high confidence (p.3). It states that the actor manipulated the Claude Code tool into attempting to infiltrate global targets. It was able to leverage AI to execute 80-90% of tactical operations independently at physically impossible request rates (p.3). The actor was reported succeeded in roughly 30 cases before Anthropic disrupted it (p.3). (In the report Anthropic said that this event was in mid-September).

    View Trusted Source

What Drives This Performance?

A breakdown of the structural factors shaping this country's score.

AI Policy

National AI strategies, laws, and oversight mechanisms that establish formal governance structures.

Contribution to overall score

63%

Civil society engagement

Participation of academia, advocacy groups, and non-state actors in shaping governance and ensuring accountability.

Contribution to overall score

7%

Enabling conditions

Institutional capacity, rule of law, digital readiness, and labour protections that enable effective governance.

Contribution to overall score

30%

  • Strongest context signal: Gender Inclusion in Connectivity (91.9)
  • Cybersecurity (91.3)
  • Population Digital Readiness (79.7)

Evidence for China

Loading evidence explorer…