The integration of Artificial Intelligence in Education (IAED) has transformed learning and teaching methods through adaptive learning systems, intelligent tutors, and personalized assessments. However, this evolution raises concerns about the transparency and understanding of decisions made by these systems, highlighting the importance of adopting Explainable AI in education. This work presents an innovative approach using local interpretability methods to analyze learner performance. The local interpretable model-agnostic explanations (LIME) technique is employed to provide specific explanations for individual predictions, enhancing understanding of factors influencing student performance. Results show that LIME captures complex interactions and interrelationships among various factors contextually. This approach improves educators’ ability to interpret model predictions, identify dropout risks, and implement more effective pedagogical interventions. Furthermore, the proposed method is innovative by integrating data augmentation techniques like SMOTE to balance classes and enhance model robustness. Through cross-validation, the performances of different models are rigorously compared, allowing for the selection of the best models based on their accuracy and generalization ability to new data. By leveraging local explanations from LIME and advanced techniques such as SMOTE and cross-validation, this work represents a novel and significant contribution to the ethical and responsible use of AI in education. This approach reinforces transparency, accountability, and equity, thereby promoting broader and more confident adoption of these technologies by educators and learners.