Methodology
How the Rankings are built and checked.
Every indicator, its source, its transformation and its weight — documented, versioned, and open to challenge.
Scoring
What a score is and is not.
Each pillar is published on a common 0–100 scale by the source framework. The six pillar scores are averaged with equal weight to produce the overall score. There is no discretionary adjustment at any stage, and no analyst may override a computed value.
Equal pillar weighting is a deliberate choice. It means a government cannot compensate for absent safeguards with abundant compute, and it means the Rankings do not encode a view about which capability matters most. Readers who disagree can re-weight the pillars themselves — the pillar scores are released alongside the overall score for exactly that reason.
Where a government has been unable to supply evidence for an indicator, the indicator is recorded as absent rather than estimated or imputed from a regional average. Absent indicators are shown on the country profile and excluded from the pillar mean, and the count of absences is published with the score. A published gap is preferred to an inferred value.
Synthesis from public data
The Rankings are a synthesis of publicly available material — official strategies, budget documents, regulatory instruments, procurement records, statistical releases and peer-reviewed research — normalized to a common 0–100 scale before aggregation. They do not commission primary surveys of the public. In this edition the pillar inputs come from one licensed upstream framework; broadening that base is the principal development objective for the next edition, and the source register records exactly what is in use.
Each indicator is mapped to exactly one pillar. Pillar scores are the mean of their constituent indicators, and the overall score is the mean of the six pillar scores. Because aggregation is arithmetic and the inputs are public, any published result can be reconstructed independently from the handbook.
Every edition carries the date on which its synthesis was produced. Where an underlying source is revised after that date, the change is carried into the following edition rather than applied retrospectively, so a score always corresponds to a stated point in time.
Sources and versioning
The 2026 edition rests on published pillar data from a single licensed upstream framework, recorded with its publisher, its license and its retrieval date on the data licensing page. Direct collection from governments through structured requests is planned for a later edition and is not the basis of these figures. Every source carries a retrieval date and a version identifier, so any score can be reconstructed from the record.
Corrections
Governments and researchers who dispute a score can submit evidence through the corrections process. Every accepted correction is logged publicly with its date, the indicator affected and the resulting change to the overall score and rank. Corrections are applied to the live dataset immediately and reflected in the next printing of the report.
Assessment framework Under development
Six pillars, weighted equally.
Equal weighting is a deliberate choice rather than a default. A government cannot offset absent safeguards with abundant compute capacity, so nothing here is allowed to buy its way past anything else.
Policy capacity
Whether a government can design and fund AI policy against a stated national vision — including whether commitment is backed by assigned resources rather than published intent, and whether it engages internationally on a technology that does not respect borders.
AI infrastructure
Compute capacity and the enabling technical infrastructure that government and others can rely on over the long term, together with data that is sufficient, high quality, and able to move securely within and across borders.
Governance
Whether governance principles exist, are promoted, and shape how AI is developed and adopted — protecting rights and societal interests through the government's own practices, and giving those subject to AI decisions an adaptive route to compliance and redress.
Public-sector adoption
Whether digital policy actually encourages testing AI against public-sector problems, and whether what proves effective is then scaled — evidenced in e-government delivery rather than in pilots.
Development and diffusion
Whether a domestic AI sector is mature enough to serve local demand and sustain long-term innovation, with the human capital for AI to diffuse to industry, researchers and civil society beyond the specialists.
Resilience
Whether the government is managing the social, economic and environmental strain of widespread adoption, and actively monitoring the safety and security risks that grow as development and diffusion progress.
An index is only useful if it is willing to publish a score its own funders dislike. That constraint shapes every methodological decision we make.
Discuss the 2026 findings with the research team.
Methodology questions, indicator-level data requests, briefings for policy and executive audiences, and media inquiries are handled directly by the assessment team.