The Drivers of Artificial Intelligence Adoption in the European Union: A Three-Level Digital Divide Perspective

Abstract

This study examines the macro-level factors associated with artificial intelligence (AI) adoption among enterprises in the European Union through the three-level digital divide framework. Despite the EU’s Digital Decade 2030 targets, AI integration remains limited and uneven across member states. Using 2024 Eurostat data for the EU27, the study applies OLS multiple linear regression and exploratory K-means clustering to examine the relationships between broadband infrastructure, ICT skills, enterprise digital intensity, and AI usage. Enterprise digital intensity shows the strongest positive statistical association with AI adoption, whereas the coefficients for ICT specialists and broadband are not statistically significant in the multivariate model. A leave-one-country-out analysis indicates that the positive DII association is not driven by any single member state. The cluster analysis identifies three distinct country profiles based on their digital characteristics. The findings suggest that infrastructure alone may be insufficient to account for cross-country variation in AI adoption and that broader organizational digital maturity should also be considered. Given the cross-sectional design, the results are interpreted as associations rather than causal effects.