The increasing use of the internet has led to new forms of data theft, which are considered a form of cybercrime. Phishing attacks are one method used to? deceive users into disclosing personal information, and URL phishing is the most? common type? of such attacks. This research aims to present an approach for detecting phishing websites by developing a machine learning model for phishing URL classification, together with hyperparameter tuning using the grid search method? across five algorithms, in order to obtain the model that provides the best phishing website classification performance. The evaluated algorithms include Neural Networks, ? Logistic Regression, Naive Bayes, Support Vector Machines using the SMO algorithm, and Decision Trees. The Weka software was used as the simulation and testing tool, and model performance was measured using five-fold cross validation. The results indicate that the Support Vector Machine model using the SMO algorithm achieved the best performance, with an accuracy of 95?.5?5%, precision of 95.60%, recall of 95.60%, and an F1 score of 95.50%. These results demonstrate that the developed model can accurately classify phishing URLs and can serve as an important prototype for the effective future development of automated phishing detection systems.