Malaysia’s energy landscape has evolved significantly over the past five decades, shifting from oil dependence to a more diversified mix. Yet, despite clear policy goals, the pace of transition toward cleaner energy has been slower than expected. This study uses AI-assisted simulation to forecast Malaysia’s 2025 electricity generation mix, drawing on historical data, policy targets, and economic trends. The AI model projected a mix of 45% natural gas, 30% coal, and 25% renewables. In reality, the actual mix was 44% coal, 37% gas, and only 19% renewables—highlighting a gap between ambition and implementation. The analysis explores key reasons behind this gap, including long-term coal contracts, volatile fuel prices, delayed renewable projects, and fragmented policy execution. It also examines the underperformance of combined-cycle gas plants and the slow rollout of solar initiatives due to grid and financing challenges. External shocks like the COVID-19 pandemic and political transitions further disrupted progress. By comparing simulated forecasts with actual outcomes, the study reveals how institutional inertia and financial constraints continue to hinder Malaysia’s energy transition. It recommends practical steps such as establishing a coal transition fund, reforming energy subsidies, integrating AI into planning, and strengthening regional energy cooperation. These measures aim to help Malaysia build a more resilient, affordable, and sustainable energy system aligned with its climate goals. This study contributes to the energy transition literature by demonstrating how AI-assisted modelling can reveal implementation gaps in national energy planning, offering actionable insights for policy reform and system resilience.