The role of artificial intelligence in energy systems development: A comparative analysis of national strategies and future directions

Artificial intelligence is expanding rapidly across energy systems, offering tools for renewable integration, forecasting, grid reliability, and demand management while also increasing electricity and infrastructure demand. Yet evidence of its energy-system benefits remains fragmented across national contexts and is often synthesised without accounting for differences in evidence quality or governance. This study addresses that gap through a systematic review of 146 peer-reviewed studies and authoritative reports published from 2010 to 2026, supplemented by earlier foundational methodological sources dating back to 1994. It compares the United States, India, South Korea, and Sweden, selected to represent contrasting market-led, emerging-economy, state-directed, and mature low-carbon contexts. The analysis combines an evidence-tiered comparison across 14 energy-system indicators, a city-level deployment index covering ten urban centres, and a four-pillar governance framework. Reported studies indicate improvements in renewable-energy forecasting performance, while CNN-based and explainable machine-learning approaches have been applied to power-system fault detection; however, country-specific quantitative validation remains uneven. Northern Virginia/DC ranks first with an unrounded AECDI score of 83.00, followed closely by Stockholm at 82.95; both scores round to 83 for presentation, but they achieve these results through different pathways: infrastructure scale and grid deployment in Northern Virginia/DC, and low-carbon electricity integration and institutional maturity in Stockholm. Across the four cases, technical performance and investment alone do not ensure sustainable outcomes; grid carbon intensity, institutional capacity, cybersecurity, explainability, and equity shape how benefits are realised. The study contributes a transparent and replicable approach that links evidence quality, sub-national deployment patterns, and measurable governance priorities, providing practical guidance for more effective and equitable integration of artificial intelligence into energy systems.