Artificial Intelligence and Multidimensional Circular Economy Performance in the EU-27: Panel Evidence

Abstract

Background and Purpose: This study examines the relationship between artificial intelligence (AI) and circular economy (CE) performance in the current European Union (EU-27) countries over the period 2001–2024.

Methods: AI-related capacity and CE performance are measured using multidimensional indices constructed through time-specific heterogeneous factor analysis. The AI index captures national AI-related readiness and adoption capacity rather than the direct intensity of AI deployment. The empirical analysis employs two-way fixed effects with Driscoll-Kraay standard errors and is complemented by lagged AI specifications, Common correlated effects mean group estimation, and Old/New EU heterogeneity analysis.

Results: The results show a positive and statistically significant association with AI-CE performance. This association remains consistent after accounting for cross-sectional dependence, serial correlation, heteroskedasticity, unobserved country heterogeneity, common time shocks, and heterogeneous country responses.

Conclusion: The heterogeneity analysis provides suggestive evidence that the AI-CE association may be stronger in Old EU countries, while the corresponding association in New EU countries is weak and statistically insignificant. These findings suggest that AI is more likely to translate into CE outcomes when supported by mature innovation systems, environmental governance, digital infrastructure, and institutional readiness.