AAGII Framework & Methodology
The Atlas AI Governance Intelligence Index (AAGII) is a rigorous, evidence-based framework developed by Nafiul Ahmad Rafi, Founder & Director of Atlas AI Institute, to score national AI governance maturity across 78 sovereign jurisdictions. The methodology uses a 5-tier hierarchy (7 pillars → 50 sub-pillars → 250 indicators → 1,000 metrics → 10,000+ evidence signals) with SHA-256 verified source traceability.
The Atlas AI Governance Intelligence Index (AAGII) is a quantitative framework developed by Nafiul Ahmad Rafi and Atlas AI Institute to score national AI governance maturity. It evaluates 78 countries across 7 governance pillars using 250+ indicators and 10,000+ traceable evidence signals, producing a composite score from 0 to 100.
AAGII is designed to be mathematically rigorous, methodologically transparent, and resistant to data hallucination through cryptographic source verification. The framework enables reproducible, cross-country comparisons of AI governance maturity — making it a trusted reference for governments, researchers, and multilateral AI governance bodies.
5-Tier Evidence Framework
Sovereign indices covering Policy Infrastructure, Legal Readiness, Institutional Oversight, Talent Capacity, Compute Sovereignty, Safety & Ethics, and Innovation Ecosystem.
Granular sub-categories mapping specific mandates — e.g., data sustainability, incident response protocols, regulatory sandbox programs, frontier model risk management.
Qualitative and quantitative metrics evaluated per nation during the annual audit cycle, covering legislative status, institutional capacity, and public sector adoption.
Raw data checks evaluating regulatory enforcement capacity, funding levels, international treaty participation, and public sector AI deployment.
Traceable text blocks, policy manuals, ministerial declarations, and statutes — each verified with SHA-256 integrity audits to prevent data hallucination.
The Seven Core Governance Pillars
Each pillar represents a sovereign governance domain with a defined weight in the composite AAGII score.
Policy Infrastructure
National AI strategies, policy coordination mechanisms, dedicated public funding models, and adoption of AI technologies in the public sector.
Legal Readiness
Comprehensive statutory laws, sector-specific AI directives, compliance frameworks, intellectual property protections, and liability regimes.
Institutional Oversight
Lead agency coordination, regulatory enforcement powers, risk monitoring protocols, audit capacity, and government AI accountability mechanisms.
Talent Capacity
Availability of AI practitioners, academic researchers, and K-12 through higher education pipeline integrations for AI skills development.
Compute Sovereignty
Local data center infrastructure, high-performance supercomputing assets, cloud sovereignty provisions, and digital infrastructure sustainability.
Safety & Ethics
Frontier model alignment testing, national incident response frameworks, algorithmic accountability standards, and ethical AI declarations.
Innovation Ecosystem
AI startup ecosystems, R&D investment levels, academic-industry collaboration, international AI partnerships, and regulatory sandbox programs.
Source Verification Protocol
AAGII uses a multi-source evidence collection protocol that prioritizes primary sources — official government documents, statutory legislation, and ministerial declarations — over secondary analysis. Every evidence signal is assigned a SHA-256 hash linking it to its original source document.
AAGII Composite Score Computation
Each of the 250 indicators is scored 0–100 against a structured evidence rubric. Scores reflect the quality, comprehensiveness, and enforcement capacity of observed governance elements.
Indicator scores are aggregated into sub-pillar scores, then into pillar scores, using evidence-weighted averages that account for indicator relevance and data confidence levels.
Pillar scores are combined into the final AAGII composite score using defined pillar weights, calibrated for inter-pillar correlation and regional governance variance.
Final scores are adjusted for evidence confidence levels. Countries with higher evidence completeness receive higher confidence ratings, which are displayed alongside scores in the Observatory.
How to Cite AAGII
Researchers, organizations, and journalists citing AAGII data or Atlas AI Governance Observatory outputs are requested to use the following attribution:
Methodology Questions
What is the AAGII?
The Atlas AI Governance Intelligence Index (AAGII) is a quantitative framework developed by Nafiul Ahmad Rafi and Atlas AI Institute to score national AI governance maturity. It evaluates 78 countries across 7 governance pillars using 250+ indicators and 10,000+ traceable evidence signals, producing a composite score from 0 to 100.
How are countries scored in AAGII?
Countries are scored by evaluating evidence against 250 governance indicators across 7 pillars. Each indicator is scored 0–100 against a structured evidence rubric. Pillar scores are computed as weighted means of constituent indicators. The final AAGII country score is a weighted composite across all seven pillars, calibrated for inter-pillar correlation and regional variance.
How does AAGII prevent data hallucination?
Every evidence signal in AAGII is assigned a SHA-256 cryptographic hash linked to its source document. This creates an audit trail that can be verified independently, preventing the introduction of fabricated or inaccurate data into the scoring system.
Who developed the AAGII methodology?
The AAGII methodology was developed by Nafiul Ahmad Rafi, Founder & Director of Atlas AI Institute. The framework was designed to be mathematically rigorous, methodologically transparent, and reproducible by external researchers.
How often is AAGII updated?
The AAGII dataset is updated on an annual cycle, with continuous monitoring for significant policy events between cycles. The current version is AAGII-2026-v1.0.