Linking clean energy consumption to environmental sustainability in Africa: Puzzling evidence from new artificial intelligence methods

Abstract The literature is inundated with studies emphasizing the importance of renewable energy consumption (REN) and sustainable resource management as a panacea to the threats of climate change, which include loss of species, food insecurity, extreme weather, sea level rise, and security risks. The current study examines the dynamic relationships in spatial and time domains between REC, natural resource rents (NR), and ecological footprint (EF) from 1990 to 2022, using the novel spatiotemporal intelligent regression models, specifically the Geographically Neural Network Weighted Regression (GNNWR) model for sub‐Saharan Africa (SSA) countries. This novel method introduced by Yin et al. (2024) integrates a spatiotemporal weighting framework with neural network architectures. The GNNWR results suggest that REN, NR, and urbanization are significant supportive elements of ecological sustainability in SSA. Conversely, non‐renewable energy consumption (NRE), economic growth, and globalization hurt the environment by upsurging the EF. The causality results display a bidirectional causality between REN and EF, NRE and EF, and economic growth and EF. However, unidirectional causality flows from NR and globalization to EF. In line with the results, policymakers in SSA will need to redefine their energy consumption portfolio. Efforts must be made to encourage energy transition by raising adequate awareness of the availability and benefits of consuming clean energies such as solar, tidal, geothermal, wind, and hydropower.