Policy pathways to renewable energy affordability: Machine learning evidence on artificial intelligence, carbon pricing, and green finance in advanced economies

The global transition toward carbon neutrality highlights the importance of ensuring renewable energy affordability in advanced economies. This study examines how artificial intelligence (AI), carbon pricing coverage (COETS), green finance (GF), renewable energy technological innovation (RETI), renewable energy share (REN), and their interaction (RETI × GF) influence renewable energy costs in G7 countries. Renewable energy affordability is proxied by the Levelized Cost of Energy Index (LCOEI) for solar photovoltaic and onshore wind technologies. To capture nonlinearities and address high-dimensional policy interactions, the analysis applies machine learning–based regularization techniques, including Standard LASSO, Adaptive LASSO, ElasticNet, and Double Machine Learning estimators for causal inference. The findings show that AI adoption, renewable energy expansion, and broader carbon pricing coverage significantly reduce renewable energy costs. In contrast, RETI alone initially increases LCOEI, reflecting early-stage capital intensity, while the RETI × GF interaction strongly lowers costs by facilitating commercialization through financial support. Robustness checks using Bayesian Model Averaging, Driscoll–Kraay, and Panel-Corrected Standard Errors confirm the stability of results. The study recommends strengthening AI-enabled grid management, expanding green finance mechanisms, and maintaining effective carbon pricing to accelerate affordable clean energy transitions consistent with SDGs 7 and 13.