Predicting customer attrition is vital in the telecommunications industry, where customer retention directly affects financial outcomes. This study evaluates six algorithms: Logistic Regression, Random Forest, Naive Bayes, K-Nearest Neighbors, XGBoost, and Support Vector Machine. Their performance was assessed using metrics such as accuracy, recall, and AUC-ROC. While Naive Bayes performed exceptionally well in identifying a greater number of at-risk customers, albeit with less precision, Logistic Regression and Random Forest were notable for their dependability. The outcomes of the other models were more inconsistent. These results point to areas where predictive tools can be improved and customer retention tactics can be reinforced.