Arab AI Governance Index 2025
Arab states are adopting national AI strategies. Dedicated AI legislation is still taking shape across the region.
The Arab AI Governance Lab studies how the 22 member states of the Arab League govern artificial intelligence. It publishes an open dataset on national strategies, regulators, data protection, readiness, research output, infrastructure, human capital and cybersecurity, and analyses why states with similar resources govern AI so differently. Its principal outputs are the Arab AI Governance Index, a comparative measure of regulatory maturity, readiness, digital government and cybersecurity across the 22 states, and the annual Arab AI Governance Report, which sets out the evidence, the gaps and the policy priorities for each state and for the region.
Filter by area
Regulatory maturity in five stages, 2025 data release. Select a state to open its scores.
Boundaries follow the usage of the League of Arab States and imply no judgement on the status of any territory.
Comparison
Compare states
Choose up to four states to compare the four pillars of the index and their AI readiness trajectory since 2020.
Pillar scores (0 to 100)
AI readiness, 2020 to 2025
Readiness values follow Oxford Insights; the 2025 method revision affects comparisons with earlier years.
Areas of work
Eight areas of analysis, each with its headline evidence and live data. Select an area.
Ten national strategies and nine dedicated bodies; a standalone AI law is still to come.
AI-relevant regulatory instruments
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Saudi Arabia moved from rank 52 to rank 6 on the UN e-government index in six years.
AI readiness score, 2025
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Scientific output and regulatory capacity do not develop in step.
AI publications, 2023
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Connectivity has expanded faster than the frameworks that govern AI.
Growth in internet use, 2018 to 2023 (percentage points)
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Five states build Arabic language models; seventeen use systems built elsewhere.
International AI frameworks referenced
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
States classified on authority structure, regulatory approach, enforcement, sovereignty posture and alignment.
Dedicated AI governance bodies
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Human capital fell in half of the states with data between 2018 and 2024.
Human Capital Index, 2024
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Five states score above 94 on the ITU index; eight score below 36.
ITU Global Cybersecurity Index
2025 release, doi:10.7910/DVN/MOVIFAOpen this area
Arab AI Governance Index 2025
Four pillars on a 0 to 100 scale with equal weights: regulatory maturity (Lab coding), AI readiness (Oxford Insights), digital government (UN EGDI) and cybersecurity (ITU GCI). The index is descriptive, not evaluative: scores summarise published frameworks and third-party indicators and imply no judgement on national policy choices.
| Position | State | Area | Stage | Index | Regulation | Readiness | Digital government | Cybersecurity |
|---|---|---|---|---|---|---|---|---|
| 1 | SAU | Gulf | Advanced | 91.8 | 100.0 | 71.6 | 96.0 | 99.5 |
| 2 | UAE | Gulf | Advanced | 90.8 | 100.0 | 69.9 | 95.3 | 98.1 |
| 3 | QAT | Gulf | Advanced | 83.9 | 100.0 | 58.6 | 82.4 | 94.5 |
| 4 | BHR | Gulf | Intermediate | 76.1 | 75.0 | 59.6 | 92.0 | 77.9 |
| 5 | EGY | North Africa | Intermediate | 74.1 | 75.0 | 59.1 | 67.0 | 95.5 |
| 6 | OMN | Gulf | Emerging | 72.1 | 50.0 | 56.7 | 85.8 | 96.0 |
| 7 | TUN | North Africa | Intermediate | 68.2 | 75.0 | 42.2 | 69.3 | 86.2 |
| 8 | MAR | North Africa | Emerging | 61.0 | 50.0 | 43.1 | 68.4 | 82.4 |
| 9 | KWT | Gulf | Initial | 55.5 | 25.0 | 43.8 | 78.1 | 75.1 |
| 10 | JOR | Levant and Iraq | Initial | 55.1 | 25.0 | 56.1 | 68.5 | 71.0 |
| 11 | DZA | North Africa | Initial | 40.1 | 25.0 | 42.0 | 59.6 | 34.0 |
| 12 | IRQ | Levant and Iraq | Emerging | 36.5 | 50.0 | 29.4 | 45.7 | 20.7 |
| 13 | LBN | Levant and Iraq | Nascent | 29.8 | 0.0 | 34.3 | 54.5 | 30.4 |
| 14 | LBY | North Africa | Nascent | 28.0 | 0.0 | 28.4 | 54.7 | 28.8 |
| 15 | MRT | North Africa | Initial | 23.8 | 25.0 | 27.6 | n/a | 18.9 |
| 16 | SYR | Levant and Iraq | Nascent | 20.7 | 0.0 | 21.7 | 38.9 | 22.1 |
| 17 | PSE | Levant and Iraq | Nascent | 20.2 | 0.0 | 35.5 | n/a | 25.2 |
| 18 | SDN | Red Sea, Horn and Indian Ocean | Nascent | 17.1 | 0.0 | 16.2 | n/a | 35.0 |
| 19 | YEM | Red Sea, Horn and Indian Ocean | Nascent | 12.6 | 0.0 | 14.5 | 23.2 | n/a |
| DJI | Red Sea, Horn and Indian Ocean | Nascent | Insufficient data | 0.0 | n/a | n/a | n/a | |
| SOM | Red Sea, Horn and Indian Ocean | Nascent | Insufficient data | 0.0 | n/a | n/a | n/a | |
| COM | Red Sea, Horn and Indian Ocean | Nascent | Insufficient data | 0.0 | n/a | n/a | n/a |
Three analytical insights that composite rankings obscure
Fiscal resources, scientific output and composite index scores are the conventional predictors of governance capacity. Configurational evidence indicates that none is a sufficient condition on its own.
Insight 1 of 3
Fiscal capacity is not a sufficient condition for institutionalisation
Kuwait combines a GDP per capita of US$34,076 and 99.75 per cent internet use with no national AI strategy or dedicated body coded in the 2025 release. The five other Gulf states have institutionalised both. The pattern indicates that institutional choice, not fiscal capacity, is the operative condition, and that designation of a lead authority offers a rapid route to institutionalisation.
Insight 2 of 3
Epistemic capacity can precede regulatory capacity
Iraq recorded 679 indexed AI publications in 2023, second among states with data, with research expenditure of 0.04 per cent of GDP and no data protection statute enacted so far. A research community of this scale constitutes absorptive capacity on which regulators can draw through formal advisory mechanisms.
Insight 3 of 3
Composite indicators can lag institutional change
The Oxford Insights readiness score for Lebanon rose from 35.91 in 2020 to 47.62 in 2023 during a severe economic crisis, then moved to 34.26 in 2025, in part because the index method was revised. Composite indicators should therefore be triangulated with institutional evidence drawn from primary documents.
Source: Arab AI Governance Lab Dataset, 2025 release.
Dr. Anis Ben Brik
Flagship report 2026
Nine chapters by Dr. Anis Ben Brik: where the region is unequal, whether it is converging, the gap between capacity and regulation in each state, 22 country notes with context, evidence and priorities, and nine priority gaps with recommendations.
Capacity building on AI governance
The Lab designs and delivers capacity building for ministries, regulators, data protection authorities and regional bodies: executive programmes, regulator training and institutional development, grounded in its comparative evidence and delivered in Arabic, English or French.
Open data
Thirteen linked tables cover regulation, readiness, research output, infrastructure, human capital, cybersecurity, fuzzy-set calibration and qualitative profiles. All files are deposited on Harvard Dataverse under CC BY 4.0.
Ben Brik, A., Gilbert, N., & Pycińska, M. (2025). Arab AI Governance Lab Dataset [Data set]. Harvard Dataverse. https://doi.org/10.7910/DVN/MOVIFA
The Lab
The Lab pursues three mutually reinforcing lines of work: peer-reviewed research on regulatory governance and algorithmic administration, open data infrastructure deposited on Harvard Dataverse, and capacity building for the public institutions that govern AI.
It is directed by Dr. Anis Ben Brik, with co-investigators and research assistants at four universities: Magdalena Pycińska (Jagiellonian University); Neil Gilbert (University of California, Berkeley); Jean-Patrick Villeneuve (Università della Svizzera italiana).