ARCHIVES

Recruitment systems that rely on Artificial Intelligence (AI) are becoming more common in today's hiring process, but they tend to reflect the demographic biases found in past recruitment data. This study investigates bias and fairness issues in automated hiring systems by developing a fairness-aware machine learning framework for recruitment decision analysis. The main goal of this study was to examine how demographic characteristics (gender and ethnicity) affect hiring predictions and to determine how well fairness mitigation techniques can minimize discriminatory hiring outcomes. The proposed framework uses machine learning models, such as Logistic Regression and XGBoost, to predict the suitability of candidates. The study further incorporated fairness mitigation strategies, such as ThresholdOptimizer, Reweighing, and Adversarial Debiasing, to enhance equitable decision-making. Cross-validation analysis, feature importance evaluation, group-wise fairness evaluation, and disparate impact ratio analysis were conducted to evaluate predictive performance and fairness across demographic groups. The findings of the experiments revealed that baseline machine learning models were highly predictive, but measurable differences in hiring outcomes across demographic groups existed. Feature importance analysis revealed that years of experience and sensitive demographic attributes significantly influenced the model predictions. Fairness-aware mitigation techniques achieved significant improvements in fairness metrics with competitive predictive performance. Adversarial Debiasing had the best fairness score among the evaluated approaches, whereas Reweighing offered a fair balance between fairness and accuracy. The results validate the effectiveness of fairness-aware optimization methods in minimizing demographic bias in AI-based recruitment systems while maintaining model performance. This study is a step toward creating responsible, transparent, and ethically sound AI recruitment systems for practical use in organizational Hiring processes.