Recent technical advancements and the fast expansion of diverse Internet of Things (IoT) networks and devices have produced vast quantities of data, with notable progress in information theft and network assaults aimed at different forms of information. This requires elevated levels of security and dependability to safeguard information against hackers. Artificial intelligence, especially machine learning, is an exceptionally efficient method for mitigating and safeguarding against cybersecurity threats. This study illustrates fundamental threat detection methodologies with machine learning algorithms. Logistic regression, a fundamental Bayesian classifier, a decision tree, and the K-Nearest Neighbor algorithm were trained using the (BoT-Net_IoT) dataset from Kaggle, a distinguished platform offering a variety of datasets across several domains. Multiple criteria, including precision, recall, and F1-score, were used to evaluate the algorithms. After comparison, the decision tree method surpassed the others and was therefore used in the system’s implementation. The decision tree method achieved a precision of 99%, an F1 score of 99%, a recall of 100%, and an accuracy of 99%.