The increasing threat of cyberattacks, particularly botnet activities, has led to the need for adaptive defensive strategies. This research project aims to improve botnet attack detection accuracy in IoT environments using machine learning. The methodology includes data pre-processing, feature selection, model training, and performance evaluation. The N-BaIoT dataset, captured before and after infection by BASHLITE and Mirai botnets, is used for training and assessment. Resampling techniques are applied for each attack type to address the imbalanced nature of the data. The models are trained using an ensemble technique called Voting Classifier, which combines the production of DT, KNN, and SVC models. The models’ performance is evaluated using the accuracy, precision, recall, and F1-score parameters. The results are compared with the preceding studies to analyze the viability and drawbacks of the suggested strategy. The project thus seeks to contribute to the growth and promotion of IoT security solutions and botnet mitigation by improving the general security environment.