Analysis of Gender Inequality in Artificial Intelligence-Based Recruitment Systems: A Systematic Literature Review (SLR)

Abstract

The increasing adoption of Artificial Intelligence (AI) in recruitment has raised concerns about algorithmic discrimination that may disadvantage certain groups, particularly women. This study analyzed gender inequality in AI-based recruitment systems by synthesizing evidence from both technical and ethical perspectives. A Systematic Literature Review (SLR) was conducted on studies published between 2020 and 2025, applying predefined inclusion and exclusion criteria, followed by screening, quality assessment, and thematic synthesis. The review retained 10 studies (n = 10) that met the eligibility and quality threshold. Historically imbalanced training data emerged as the most frequently reported driver of gender bias, often producing unfair screening, ranking, and selection outcomes. Fairness conclusions were found to depend strongly on how recruitment outcomes were defined and measured, and prior studies consistently called for multiple fairness metrics supported by auditing practices. The literature also identified mitigation strategies spanning data balancing, fairness-aware model evaluation, transparency and audit mechanisms, and human oversight in decision-making. Gender bias in AI-based recruitment is, at its core, a socio-technical problem that requires combined interventions across data governance, model evaluation, and organizational accountability, while research gaps remain for future empirical validation and responsible AI deployment.