This study aims to predict students’ academic success using Artificial Intelligence and Machine Learning techniques applied to the Student Performance Factors dataset. The dataset comprises 6,607 observations and 20 variables, including academic, demographic, and social characteristics. After data cleaning, handling missing categorical values, encoding categorical variables, and standardizing numerical features, a binary target variable (Passed) was constructed based on the median of the Exam_Score. Exploratory Data Analysis revealed that academic engagement factorsparticularly attendance, study hours, and previous academic results— exhibit a stronger relationship with exam success than demographic or social factors. Several classification models were evaluated, including K-Nearest Neighbors, Decision Tree, and Logistic Regression, with Logistic Regression achieving the best overall performance. The final model demonstrated strong predictive ability, achieving an accuracy of 88.7%, precision of 89.1%, recall of 91.1%, and an F1-score of 90.1%. The results confirm that academic performance can be effectively predicted using Machine Learning approaches and highlight the critical role of academic engagement indicators. These findings suggest that such models can support educational institutions in early identification of students at risk and in designing targeted interventions to improve academic outcomes.