Press "Enter" to skip to content

Defining indicators for the African AI Index report – A Framework for Measuring AI Readiness on the continent

AI is transforming the world and Africa is no exception, governments on the continent need to integrate AI into the public sector and create an environment that fosters its use by the private sector.

Learn about the 25 indicators used in the African AI Index Report a comprehensive framework for governments on the continent to assess national AI readiness.

One might ask why not use the current AI readiness tools, it is simple, they require data sets unavailable on the African continent.

We selected 25 AI readiness indicators whose data is available in all 54 countries on the African continent. Call our African AI Index a more contextual, and practical AI readiness framework African for governments.

The 25 AI Readiness Indicators are categorized into 5 thematic areas: AI Regulations, AI infrastructure, AI Skills, AI Innovation, and Government AI initiatives.

AI RegulationsAI InfrastructureAI SkillsAI InnovationGovernment AI initiatives
AI legislationElectricity access rateComputer Science graduates per yearAI start-upse-Government
Data protection and privacy legislation Internet penetration rates AI graduates per year AI research papers published per yearGovernment AI research funding
National AI StrategyCost of data bundlesExpert AI GroupsAI patents per yearAI research labs & hubs
Cybersecurity legislation Computer access per household AI job ads per year Academic-Business partnershipsOpen government data
AI regulatory bodySmartphone penetration rateComputer literacy AI start-up incubators and accelerators AI literacy in the public sector
Computers per studentFemale computer science graduatesAccess to Capital AI public-private partnerships

Outlier AI Readiness Indicators that we excluded

Using Oxford Insights 2023 Government AI Readiness Index as a benchmark find a full list of the AI readiness indicators we didn’t include and our logic for excluding them.

AI Readiness Indicator Reason for exclusion
Supercomputers You don’t need supercomputers for AI
Number of non-AI
technology unicorns
Unicorns represent a tiny proportion of overall entrepreneurial activity
Value of trade in ICT good and services
(per capita)
No direct correlation with AI readiness
Company investment in
emerging technology
Unreliable data on this indicator
Number of AI unicornsUnicorns represent a tiny proportion of overall entrepreneurial activity
Regulatory qualityDifficult to measure
National ethics framework
(Y/N)
Captured under AI legislation
AccountabilityNo direct correlation with AI readiness
Government effectivenessDifficult to measure and very subjective
Government’s
responsiveness to change
Difficult to measure
GitHub users per thousand
population
Unreliable data
Quality of engineering and
technology higher
education
Subjective datasets used for this indicator
Graduates in STEMA generic indicator that can be better captured by referencing more direct skills
Gender gap in Internet
access
Data on this indicator is not available at scale in Africa
Cost of internet-enabled
device relative to GDP per
capita
A generic indicator that can be better captured by using more specific data points.
Statistical capacityNo direct correlation with AI readiness
Data governanceCaptured under AI legislation
Adoption of emerging
technologies
Difficult to measure
5G infrastructureIndirectly captured by Internet speeds
Time spent dealing with
government regulations
Too generic

This research project happened by accident when I wrote an article on the African Union launching a Continental AI Strategy which led to another article Comparing African AI regulations and the rest as they say is history.

I will write follow-up articles to provide more details about the AI Readiness indicators, scores, weighting, and calculations for the African AI Index culminating in a full report that will be published annually.

Here’s to Africa realizing the full benefits of AI!