Machine learning discovery of regional and social disparities in electric vehicle charging reliability with GPT-5

A growing body of literature has documented that unreliable electric vehicle (EV) charging poses a major barrier to public infrastructure for climate mitigation. However, prior methods for detecting reliability have been inadequate for revealing regional and social disparities in EV charging reliability at scale. This study addresses that gap by developing a machine learning pipeline that analyzes 838,785 unstructured consumer reviews to uncover critical disparities in EV charging performance. We demonstrate that zero-shot and few-shot learning with iterative expert prompts substantially reduces training data needs, achieving new performance benchmarks for domain-aware reliability detection (F1 score: 0.97, SD: 0.02). We further show how station reliability detection can be combined with diversity indices for spatial analysis to inform economic and policy decision-making in infrastructure management. We find evidence of widespread charging reliability issues across 1,653 U.S. counties, affecting over 300 million people. Disparities in charging reliability are most pronounced in metropolitan areas and along federally designated EV-ready corridors, raising concerns about inconsistent user experiences in high-traffic zones. Current incentives prioritize deployment over reliability, leaving challenges persistent without further policy interventions. We provide a credible, evidence-based, and scalable machine learning framework to identify infrastructure risks, supporting a more reliable and equitable transition to electric mobility.