From disruption to adaptation: the nonlinear associations between AI-enabled automation and climate resilience

Climate change poses escalating systemic risks to economies worldwide, and achieving the resilience objectives embedded in the Sustainable Development Goals requires adaptive capacity that integrates technological innovation with sustainable resource management. Against this backdrop, this study examines the nonlinear relationship between artificial intelligence and climate resilience in APEC economies from 1995 to 2023, using a two-way fixed-effects model to identify nonlinearities and further examining the role of renewable energy and natural resource rents. The empirical results reveal a clear U-shaped relationship between AI and climate resilience, indicating that AI initially reduces climate resilience but enhances it after surpassing a turning point. This U-shaped pattern holds in economies with higher levels of renewable energy development, whereas in economies with lower renewable energy shares, the relationship between AI and climate resilience is not statistically significant, and no significant nonlinear pattern is identified. Furthermore, natural resource rents weaken both the strength and curvature of AI’s influence on climate resilience. By integrating AI, energy structure, and resource dependence into a unified nonlinear framework, this study extends the understanding of how technological transformation can generate both resilience challenges and adaptation opportunities and provides policy insights for promoting climate-resilient development.