We measure government commitments and country capacities towards the responsible development of AI, through a technical, social and political lens. Moving beyond the parameters of innovation and investment, the Index fills critical data gaps by employing human-rights based benchmarks, and covering continents and countries not usually found in a study of this kind.
What is responsible AI?
We define responsible AI as: The design, development, deployment and governance of AI in a way that respects and protects all human rights and upholds the principles of AI ethics through every stage of the AI lifecycle and value chain. It requires all actors involved in the national AI ecosystem to take responsibility for the human, social and environmental impacts of their decisions. The responsible design, deployment and governance of AI are proportionate to the purpose of its use and meet the technological needs of the individuals and societies it seeks to serve.
Our conceptual framework
The Index measures 19 thematic areas of responsible AI, across three dimensions.
Each thematic area assesses the performance of three different pillars of the responsible AI ecosystem: Government frameworks, government actions, and non-state actors’ initiatives.
Methodology and Scoring
Between November 2023 to February 2024, data for the Index was collected first-hand by 138 in-country researchers. Researchers completed a comprehensive expert survey containing 1862 questions (98 per thematic area) designed to ascertain conditions and actions being taken to advance responsible AI in each country surveyed. A global team of quality assessors conducted an exhaustive review of all data collected.
The inclusion of nations in the initial iteration of the Global Index on Responsible AI relied on the availability of appropriate country-specific researchers and the pre-existing ties with research institutions. The decisions about which countries were part of the Index did not involve governmental bodies.
Index scores were derived from primary data gathered across the three pillars. Each thematic area was scored on each pillar, scaled to a 0-100 range, and averaged to compute pillar scores.
The scores from the primary data are then adjusted using a coefficient built from secondary data for each country, which accounts for the effectiveness of government frameworks, government actions, and the enabling environment for non-state actors.
Each pillar is assigned a specific weight and aggregated into an overall score which, allows a comparative assessment of efforts made by countries to promote responsible AI and to rank them accordingly.
