Enhancing Transparency in Credit Decisions: Explainability for AI Models

Today, AI technology is used in the emerging industry of fintech to accelerate and refine credit card acceptance processes through automation. This study aims to illuminate how AI makes credit card approvals a black box riddle. Thus, it uses XAI (Explainable AI) methods like SHAP and LIME. Our research opens the machine learning classifier mechanism and identifies the master variables that determine credit card application approval using these plausible methods. We use SHAP and LIME algorithms to predict credit card approvals using our 5,000-record dataset with factors including Customer ID, Age, Work Experience, and Annual Income. This study suggests that future AI applications for finance will emphasize higher understanding and trust in AI decision-making, making them more ethical and transparent. This exploration aims to examine the importance of XAI techniques as common instruments in financial decision-making and encourage governments, the business sector, and academia to integrate them.