This paper presents a novel hybrid model for short-term load forecasting (STLF) that integrates multiple artificial intelligence (AI) models with Lasso regression using the stacking technique. The base learners include XgBoost, LSTM, Stacked LSTM, and Bi-LSTM, while Lasso regression serves as the metalearner. The model aims to improve the accuracy of STLF by considering factors such as temperature, rainfall, and daily electricity prices. Empirical analyses on real-world datasets from Australia and Spain show significant improvements in forecasting accuracy, with a substantial reduction in the mean absolute percentage error (MAPE) compared to existing hybrid models and individual AI models. The research highlights the efficiency of the stacking technique in enhancing STLF accuracy, suggesting potential operational efficiency benefits for the power industry. The model's effectiveness is demonstrated through its ability to accurately predict load under various conditions, including stable and unstable datasets, and its robustness in handling complex relationships and external factors. The paper also discusses the economic and power grid implications of the proposed model and suggests future directions for improvement, such as integrating more advanced hybrid models and incorporating additional variables.This paper presents a novel hybrid model for short-term load forecasting (STLF) that integrates multiple artificial intelligence (AI) models with Lasso regression using the stacking technique. The base learners include XgBoost, LSTM, Stacked LSTM, and Bi-LSTM, while Lasso regression serves as the metalearner. The model aims to improve the accuracy of STLF by considering factors such as temperature, rainfall, and daily electricity prices. Empirical analyses on real-world datasets from Australia and Spain show significant improvements in forecasting accuracy, with a substantial reduction in the mean absolute percentage error (MAPE) compared to existing hybrid models and individual AI models. The research highlights the efficiency of the stacking technique in enhancing STLF accuracy, suggesting potential operational efficiency benefits for the power industry. The model's effectiveness is demonstrated through its ability to accurately predict load under various conditions, including stable and unstable datasets, and its robustness in handling complex relationships and external factors. The paper also discusses the economic and power grid implications of the proposed model and suggests future directions for improvement, such as integrating more advanced hybrid models and incorporating additional variables.