The COVID-19 pandemic accelerated the adoption of telecommuting across industries, reshaping electricity consumption patterns by shifting loads from commercial to residential sectors. This paradigm shift presents new challenges for power system operators who must adapt forecasting methodologies to maintain grid stability, reliability, and operational efficiency. Traditional statistical methods, such as ARIMA and exponential smoothing, though robust under stationary conditions, struggle to capture the nonlinearities introduced by telecommuting behaviors. Conversely, machine learning (ML) techniques, including support vector regression (SVR), artificial neural networks (ANNs), long short-term memory (LSTM) networks, and ensemble learning, offer enhanced capacity for modeling nonlinear dynamics but require abundant high-quality data. This review synthesizes advances in statistical and ML approaches for short-term load forecasting (STLF), focusing on the impact of telecommuting at institutional and regional levels, with the Federal University of Agriculture, Zuru (FUAZ), serving as a case study. Contributions from Nigerian researchers, including Gelwasa’s works on AI-driven power system stability, are highlighted to situate the research within local and global contexts. The study concludes that hybrid models integrating ML with statistical frameworks provide superior forecasting accuracy under telecommuting-induced uncertainty. Recommendations are made for utilities and research institutions to invest in hybrid models, enhanced data acquisition, and context-specific modeling