Global Index on Responsible AI (GIRAI)

Why GIRAI Matters

Turning responsible AI committments into evidence, accountability and action.

Artificial intelligence is rapidly transforming how societies work, shaping everything from public services and education to healthcare, employment, and access to information. As governments and institutions increasingly adopt AI systems, the quality of AI governance will play a critical role in determining whether these technologies advance human wellbeing or deepen existing inequalities and harms.

While many countries have introduced AI strategies, policies, and ethical principles, responsible AI cannot be achieved through commitments alone. Effective governance requires institutions, oversight mechanisms, public participation, implementation capacity, and accessible pathways for accountability and redress. Without evidence, it is difficult to know whether these safeguards exist, how well they are functioning, and whether they are reaching the people they are intended to protect.

The Global Index on Responsible AI was established to provide this evidence. Grounded in a human rights-based approach, GIRAI measures how countries are translating responsible AI commitments into action. It assesses the policies, institutions, initiatives, and broader governance conditions that support ethical, inclusive, and rights-respecting AI.

By collecting country-level evidence from around the world, GIRAI provides a comparative view of responsible AI across diverse governance, economic, and social contexts. It helps policymakers identify gaps and priorities, enables civil society to hold institutions accountable, supports researchers and international organisations with robust evidence, and highlights emerging practices that can inform global learning.

Ultimately, GIRAI exists to answer a simple but important question: are countries building the conditions needed for AI to serve people, protect rights, and contribute to more just and inclusive societies?

SECOND EDITION

Global Index on Responsible AI — 2026 Report

Download the latest edition of the Global Index on Responsible AI, offering comparative insights into how countries govern and use artificial intelligence.

Takeaways

Key Findings

What the data from 135 countries and jurisdictions reveals about the state of AI governance worldwide.

View All Takeaways

01Governance

AI is accelerating faster than governments can govern it in the public interest

Diffusion of AI is expanding, with 53% of the global population having used generative AI tools. Yet average GIRAI scores remain low, at roughly 35 out of 100, and evidence of implementation exists in only 55% of cases where frameworks are active, falling to 45% in Global South countries.

53%

of the global population has used generative AI tools

02Global South

Responsible AI governance is expanding but binding protections remain weak

Global South countries substantially broadened responsible AI content in their national frameworks — topics covered rose from 2.5 to 4.7, an 83% increase. Yet 78% of responsible AI framework cases in these countries are non-binding, compared with 42% in Global North countries.

83%

increase in responsible AI topics covered by Global South frameworks

03Safety

AI safety is governed as a technical problem while human harms remain under-addressed

AI safety and security is one of the fastest-growing areas, but much of it focuses on technical safeguards. The Index found credible evidence of government misuse of AI in 35 of 135 countries, and only 36% of countries have frameworks addressing AI-facilitated misinformation and violence.

35

countries with credible evidence of government AI misuse

04Transparency

Governments are regulating AI transparency but not disclosing their own use of AI

Transparency and Explainability is the strongest-performing indicator, with 58% of countries having some form of framework. Yet Public Disclosure of Government Algorithmic Systems is the weakest-performing indicator.

18%

of countries require disclosure of government AI systems

05Gender

Gender is increasingly recognised but protection from gendered harms remains weak

Gender equality is gaining visibility, with 29 new countries addressing gender and AI since the 1st Edition, but only 24 of 55 countries with gender-related frameworks show evidence of implementation.

29

new countries addressing gender and AI since the 1st Edition

06Education

Future generations are prepared for the AI economy but not protected from harms

AI Literacy is one of the strongest-performing indicators, with 71 countries (53%) having a framework in place. By contrast, only 55 countries (41%) have frameworks addressing Children's Rights in AI, and only 27 show evidence of implementation.

53%

of countries have an AI literacy framework in place

07Environment

AI's environmental footprint remains a blind spot in responsible AI governance

Only 27% of countries have frameworks addressing AI's environmental effects, and 83% of those frameworks are non-binding. Very few governments require disclosure of AI's energy use, water use, or environmental impact.

27%

of countries have frameworks on AI's environmental effects

08Inclusion

Governments recognise the need for local-language AI but do not require developers to deliver it

Governments are investing in local-language technologies, with 52 countries (39%) showing government-led initiatives. Only 47 countries (35%) have frameworks addressing Cultural and Linguistic Diversity, and few require developers to use diverse datasets.

35%

of countries have a cultural and linguistic diversity framework

09Labour

Governments are investing in AI skills but neglecting workers' rights

Labour protection frameworks exist in only 39 countries (29%), compared with 72 countries (53%) with frameworks on reskilling and upskilling. Few countries address workers' rights to organise and collectively bargain.

29%

of countries have a labour protection framework

10Equity

Global AI governance is fragmenting before a shared floor of protection is established

Average GIRAI scores range from 55 in Global North to 27 in Global South countries. Only 73 of 135 countries (54%) have adopted a national AI policy, and just 36 countries (27%) have operational mechanisms for civil society participation.

54%

of countries have adopted a national AI policy or equivalent

Evidence Explorer

Explore the Evidence Behind the Index

Access the laws, policies, strategies, institutional actions, and public documents that inform GIRAI scores across countries and regions.

Examine the Evidence

1,872Documents reviewed

135Countries assessed

38Indicators covered

7Global regions

Every score in GIRAI is grounded in publicly verifiable evidence, reviewed through a structured research and validation process.


1	Norway	75.3	Europe	1st of 32	1st of 45 – High income
2	Italy	72.7	Europe	2nd of 32	2nd of 45 – High income
3	Ireland	71.4	Europe	3rd of 32	3rd of 45 – High income
4	France	70.3	Europe	4th of 32	4th of 45 – High income
5	Netherlands	69.5	Europe	5th of 32	5th of 45 – High income
6	Germany	69.0	Europe	6th of 32	6th of 45 – High income
7	United Kingdom of Great Britain and Northern Ireland	67.3	Europe	7th of 32	7th of 45 – High income
8	Slovenia	66.1	Europe	8th of 32	8th of 45 – High income
9	Latvia	65.2	Europe	9th of 32	9th of 45 – High income
10	Estonia	64.9	Europe	10th of 32	10th of 45 – High income
11	Brazil	63.3	South and Central America	1st of 14	1st of 35 – Upper middle income
12	Spain	63.3	Europe	11th of 32	11th of 45 – High income
13	Greece	62.4	Europe	12th of 32	12th of 45 – High income
14	Chile	61.9	South and Central America	2nd of 14	13th of 45 – High income
15	Bulgaria	61.6	Europe	13th of 32	14th of 45 – High income

Indicator	Norway	Italy	Europe
Inclusion and Diversity7
APGender Equality	72.5	62.7	51.1
APChildren's Rights	80.3	73.9	59.0
APCultural and Linguistic Diversity	66.7	72.3	41.8
CSOCivil Society Engagement in Inclusion and Diversity	38.5	15.4	21.4
ECEgalitarian Democracy	100.0	76.8	67.6
ECDevice Affordability	89.6	73.6	76.1
ECGender Gap in Mobile Internet	93.9	100.0	95.3
Ethics and Sustainability7
APFairness and Non-discrimination	73.9	83.0	60.1
APTransparency and Explainability	73.9	83.0	64.5
APHuman Oversight and Determination	73.9	74.5	57.8
APEnvironmental Impact	70.1	80.7	28.5
CSOCivil Society Engagement in Ethics and Sustainability	15.4	76.9	30.8
ECLow-Carbon Energy Share	98.6	50.2	64.7
ECRule of Law	100.0	68.0	69.2
Labour and Skills6
APLabour Protections	73.2	73.2	53.9
APReskilling/Upskilling Initiatives	66.7	73.1	56.0
APAI Literacy	66.0	73.2	52.3
CSOCivil Society Engagement in Labour and Skills	51.3	65.6	30.1
ECLabour Rights & Compliance	100.0	47.6	62.9
ECSkills & Literacy	92.3	94.1	87.4
Trust and Safety10
APSafety and Security	74.5	83.6	60.7
APAccess to Redress and Remedy	73.9	73.9	55.0
APImpact Assessments	73.2	73.2	60.4
APAI-facilitated Misinformation and Violence	79.5	88.6	58.4
CSOCivil Society Engagement in Trust and Safety	39.6	30.8	24.0
ECData Protection and Privacy	66.8	100.0	84.2
ECData Sharing and Access	100.0	93.8	87.1
ECConsumer Protection	97.7	100.0	91.6
ECCybersecurity	96.9	100.0	88.6
ECGlobal Peace	82.3	81.5	79.4
AI Use in Public Service7
APPublic Sector Skills Development	69.3	72.9	42.0
APPublic Disclosure of Government Algorithmic Systems	63.8	62.9	21.5
APPublic Procurement	64.8	63.9	25.3
CSOGovernment Mechanisms for CSO Inclusion in AI Policy and Governance	70.0	70.0	43.8
ECCivil Society Accountability	96.3	80.0	75.5
ECPublic Service Delivery	79.8	90.1	86.3
ECAccess to Public Information	100.0	64.6	62.9
Score adjustment
URAIUnacceptable Risk AI Systems	No penalty	No penalty	×0.98Regional average

Three Pillars of Responsible AI Governance

To provide a comprehensive picture, the indicators for the Global Index on Responsible AI are organised into three pillars.

AI Policy

Assesses the presence, scope, and implementation of public-sector policies, institutions, and actions guiding responsible AI use.

Civil society engagement

Captures the extent to which civil society, academia, and non-state actors contribute to shaping and overseeing responsible AI.

Enabling conditions

Reflects the broader national conditions, capacities, and environments that influence responsible AI development and governance.

Understanding the Limits of Measurement

What this Index can — and cannot — capture.

What The Index Can Do

The Index helps users understand the global landscape of responsible AI governance. By comparing countries and regions across common indicators, it reveals where progress is being made, where challenges remain, and how governance approaches differ across contexts.

What The Index Cannot Do

Responsible AI is contextual, political, and contested. Scores cannot capture the lived realities of those affected by AI systems, the nuances of implementation, or the complex social and ethical trade-offs that arise in practice.

How GIRAI Responds

GIRAI complements quantitative scores with qualitative insights, transparent methodology, and contextual interpretation. We recognise that measurement is a tool — useful for comparison, but never a substitute for critical judgment. Rather, it is a resource needed for deepening our understanding of how AI systems shape people's lived realities.

Our Impact on Responsible AI Governance

How GIRAI is helping governments, researchers, civil society, and institutions understand and strengthen responsible AI governance worldwide.

01 — GIRAI

International Organisation

GIRAI provides clear benchmarks and comparative insights that help policymakers identify gaps, track progress, and design stronger governance frameworks for artificial intelligence.

02Global Policy makers

03Media Institutions

GIRAI in Action

Trusted by organisations worldwide

Governments, researchers, and institutions that rely on GIRAI data and reports.

Igniting global action on responsible AI, with local evidence.