The efficient identification of malicious URLs has become crucial due to their growing hazard to individuals, companies, and digital infrastructure. This study evaluated multiple machine learning algorithms for their ability to predict and identify dangerous URLs. The research focused on the Random Forest Classifier since it outperformed rival models in binary and multi-class classification tasks. With 98.9% accuracy in binary classification, the Random Forest Classifier performed well. This shows the classifier can identify safe and hazardous URLs. The system's precision of 98.8%, F1 score of 99.3%, true positive rate of 99.7%, and true negative rate of 95.6 demonstrate its dependability. Multi-class classification accuracy was 97.0%, and precision, recall, and F1 scores were good again for the Random Forest Classifier. This research provides practical tips for enhancing web security and shows how transparent AI models and interdisciplinary teamwork may solve complicated cybersecurity problems. This research has made a significant contribution to the body of known information, and its significance lies in the fact that it provides both benefits.