This study explores the application of Explainable Artificial Intelligence (XAI) techniques in credit risk assessment to enhance financial decision-making transparency and effectiveness. We developed a Multilayer Perceptron (MLP) classifier and trained it on the Home Credit Default Risk dataset to predict credit risk. The model achieved an accuracy of 93.93%. We employed SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-agnostic Explanations) analyses to interpret the model's decisions, revealing key factors influencing credit risk predictions such as goods price, credit amount, and income. Based on these insights, we created an enhanced MLP classifier focusing on the most important features identified by SHAP analysis. This refined model demonstrated improved performance metrics, including increased accuracy to 95.23%. The integration of machine learning with XAI techniques not only improved predictive accuracy but also provided crucial transparency for regulatory compliance and stakeholder confidence in the financial sector. Our approach offers a practical framework for financial institutions to leverage advanced AI models while maintaining interpretability and accountability in their decision-making processes.