AI Governance Observatory
A live-updating benchmark that tracks how prepared national governments are to govern artificial intelligence, measuring policy maturity, institutional capacity, and data infrastructure across countries — with a dedicated focus on the Global South.
An interactive map view for browsing governance and readiness scores by country and region, available on the live platform. Open Global Map →
A browsable index of the legal texts, national strategies, and official sources underlying each country's score, available on the live platform. Open Evidence Explorer →
A scenario tool for exploring how policy or institutional changes could shift a country's readiness score, available on the live platform. Open Simulation Lab →
An AI-assisted research assistant for questions about national AI governance data, available on the live platform. Open AI Policy Copilot →
What the Observatory Measures
The Observatory scores each country against five weighted indicators designed to capture the practical readiness of a government to govern AI responsibly, not just the existence of a stated ambition.
- 01
Governance Score
Measures whether binding AI-related laws, regulations, or executive orders are in force, and how enforceable they are in practice.
- 02
Readiness Score
A composite of the remaining four indicators, reflecting a country's overall preparedness to manage AI risks and opportunities.
- 03
Policy Depth
Assesses whether national AI strategies address safety, bias, procurement, and labor impacts, or remain high-level statements of intent.
- 04
Institutional Capacity
Evaluates whether a designated regulatory body, oversight unit, or ministry mandate exists with the authority to act on AI issues.
- 05
Data Infrastructure
Tracks the availability of public data governance frameworks, national data protection law, and open government data standards.
Live Index Preview
A preview of how country entries will appear once the Observatory dataset goes live. Every score below is a placeholder used for design and layout purposes only.
| Country | Governance Score | Readiness Score | Status |
|---|---|---|---|
| Bangladesh | 62 / 100 | 54 / 100 | In Progress |
| Kenya | 58 / 100 | 61 / 100 | Active |
| Indonesia | 67 / 100 | 59 / 100 | Active |
| Nigeria | 49 / 100 | 45 / 100 | Under Review |
| Brazil | 71 / 100 | 68 / 100 | Active |
| Vietnam | 55 / 100 | 50 / 100 | In Progress |
Methodology Summary
Each country's score is built from publicly verifiable evidence — legal texts, official strategy documents, government gazettes, and independent policy trackers — scored against a shared rubric to allow cross-country comparison.
Researchers assign a score of 0–100 to each of the five core indicators using a documented rubric. The four sub-indicator scores are combined into a single Readiness Score using a fixed weighting formula, described in full in the published methodology. Every score is paired with a source citation and a last-reviewed date so that users can trace how a number was derived.
Read Full Methodology →- Identify primary sources: laws, strategies, official registries.
- Score each indicator against a documented rubric.
- Apply peer review by a second regional analyst.
- Publish with citations and a last-reviewed date.
- Re-score on a fixed update cycle or after major policy change.
How to Use This Data
The Observatory is designed to be cited, compared, and built upon. Guidance below helps you interpret scores responsibly and reference them correctly in your own work.
Interpreting scores
- Scores measure documented policy and institutional readiness, not AI adoption or economic output.
- A lower score does not imply the absence of any AI activity in a country, only limited formal governance infrastructure at the time of assessment.
- Compare countries within similar income or regional groupings for the most meaningful reading.
- Always check the last-reviewed date attached to each entry before citing.
Using the dataset responsibly
- Cite the Observatory and the specific data release date, not just "Atlas AI Institute."
- Do not present illustrative or placeholder figures as final published data.
- Link back to the source country entry when referencing a specific score.
- Contact the research team for bulk or programmatic data access.
Built with researchers, civil society, and policymakers
Read our published research, join the coalition shaping the next data release, or get in touch with the team directly.
FAQ
The AI Governance Observatory is a live-updating benchmark from Atlas AI Institute that tracks AI governance readiness and policy maturity across countries, with a particular focus on Global South nations that are underrepresented in existing global indices.
Each country receives a Governance Score and a Readiness Score derived from five weighted indicators — governance, policy depth, institutional capacity, and data infrastructure — scored against a documented rubric using publicly verifiable sources. Full weighting details are published in the Methodology.
The Observatory is designed to update on a fixed research cycle, with additional updates triggered by significant policy changes in a tracked country. Update cadence will be confirmed and published alongside the first live data release.
Yes. Once live, Observatory data may be cited using the suggested citation format shown above, provided illustrative or placeholder figures are not represented as final published data. Contact the research team for bulk or programmatic access.