Cross-country gaps in artificial intelligence: Factors explaining digital inequality

Abstract

This study examines cross-country differences in Artificial Intelligence (AI) development, emphasizing the role of the digital divide. First, countries are classified into advanced, emerging, and lagging groups using cluster analysis. Then, a probabilistic model assesses how socio-economic factors such as GDP, Human Development Index (HDI), business density, and skilled labor unemployment, influence the likelihood of countries belonging to the AI emerging/advanced cluster. Results show that higher GDP per capita, skilled labor unemployment and HDI increase the likelihood of belonging to the AI emerging/advanced group. AI tends to deepen the pre-existing digital or connectivity divide. The findings underscore the need for policies and coordinated strategies that promote AI adoption and address ICT appropriation disparities, addressing structural socio-economic constraints. Advancing from lagging to an emerging/advanced position requires deep transformations in digital infrastructure, human capital and access to quality data.