International Research Journal of Education and Technology

The integration of Artificial Intelligence (AI) has revolutionized decision- making in sectors ranging from healthcare and finance to cybersecurity. However, the rise of sophisticated "black-box" models— notably deep learning—has created a transparency gap, where high predictive accuracy comes at the cost of interpretability. Explainable Artificial Intelligence (XAI) serves as a vital bridge, providing human-centric insights into how models reach specific conclusions. This paper surveys the XAI landscape, categorizing core methodologies like LIME and SHAP, and exploring how transparency fosters user confidence.

Keywords: Explainable AI, Interpretability, Machine Learning Transparency,

Human-Centered AI