Predicting loan approval is an essential task for banks and financial institutions, as it entails assessing the risk and profitability of lending money to potential borrowers. Loan approval processes in financial institutions are often complex and time-consuming, relying on manual assessments that can be biased and inconsistent. These difficulties make it more difficult to properly evaluate risks and guarantee ethical lending practices. To address these issues, this research proposes an approach rooted in machine learning to anticipate the approval status of loans for applicants based on their personal and financial characteristics. The research depends on a dataset consisting of 5000 loan applicants with 14 attributes, acquired from Kaggle. We assess the model using three classification techniques such as decision tree, k-nearest neighbors (KNN), and support vector machine (SVM). Metrics like accuracy, precision, recall, and F1-score were used to assess the performance of the models. The SVM model demonstrates an accuracy of 96%, while the decision tree achieves 92% and KNN attains 86%. According to these findings, we conclude that using the SVM model is a reliable and effective method for predicting loan approval status.