AI 채용시스템과 공정성 인식: 최신 연구 리뷰와 향후 연구 과제

Purpose - This study aims to comprehensively review empirical research on fairness perceptions in AI-based recruitment systems, by identifying key influencing factors and by examining the trends in fairness-related perceptions across countries, methodologies, and time periods. Design/methodology/approach - This study summarizes and analyzes 38 highly-impacted empirical studies published between 2015 and 2025 through a structured literature review approach. These studies are classified by country, research design (cross-sectional or longitudinal), direction of fairness perception (positive/negative/mixed), and key variables. Findings - The findings show that fairness perceptions toward AI recruitment systems were highly heterogeneous, with both positive and negative evaluations depending on the system’s explainability, transparency, and decision-maker types (AI vs. human). While some studies report that AI systems increase fairness by reducing human bias, others suggest that perceived fairness was decreased due to a lack of transparency, explainability, or individual tailoring. evertheless, the majority of studies have some limitations in adopting a simple model, using the cross-sectional data, or conducting small-scale experiments or surveys, thereby limiting generalizability and causality. Research implications or Originality - This study contributes to the AI-based recruitment system and fairness literature by systematically organizing previous research. Based on previous studies, this study identifies three major limitations: lack of theoretical integration, limited methodological rigor (e.g. cross-sectional designs), and insufficient attention to practical applicability. Furthermore, following on the limitations of previous research, this study proposes theoretic integration, methodological elaboration, more consideration of contextual factors, and increasing practical implications as the directions of future research.