Artificial Intelligence Adoption and Circular Economy Performance: Global Insights Into Resource Circularity Dynamics

ABSTRACT

Global economic development remains largely unsustainable under the dominant linear “take‐make‐dispose” production model. This model accelerates resource depletion, waste generation, and environmental degradation and undermines ecological stability. The circular economy (CE) offers a systemic alternative by improving resource efficiency, reuse, and closed‐loop production processes. However, prior empirical studies have focused mainly on traditional macroeconomic determinants of circularity, often overlooking the role of emerging technologies in facilitating national CE transitions. This study investigates how artificial intelligence (AI) adoption shapes country‐level CE performance, with particular attention to the complementary roles of digital infrastructure and research and development (R&D) intensity. Taken together, the resource‐based view, contingency theory, and ecological modernization theory conceptualize AI as a strategic technological resource. Using panel data for 67 countries from 2015 to 2023, we construct a comprehensive entropy‐weighted AI adoption index and a PCA‐based CE index. The analysis is based on the common correlated mean group (CCEMG) and augmented mean group (AMG) estimators, and the robustness of the outcomes is verified through the Driscoll–Kraay estimator, spatial autoregressive (SAR), and spatial error (SEM) models. The system generalized method of moments (Sys‐GMM) is further employed to address potential endogeneity. The results indicate that a 10% increase in AI adoption leads to a 0.98% increase in CE performance globally. Digital infrastructure serves as a key transmission channel and R&D intensity strengthens this effect. Renewable energy consumption, government waste‐recycling expenditure, and economic growth significantly enhance CE performance. In contrast, natural resource rents and carbon emissions inversely affect CE performance. This study suggests that AI is a necessary but not sufficient condition for advanced circularity. Effective CE transitions require investments in digital infrastructure and R&D intensity to decouple economic growth from environmental degradation and advance sustainable, low‐carbon economies.