Skip to content
Review Open access

The Impact of Artificial Intelligence on Recruitment Decision-Making: An Integrative Critical Review of Opportunities, Risks, And Human Oversight

2026 · Research and Analysis Journal · 0 citations · 28 references

Abstract

Artificial intelligence (AI) is moving recruitment from a largely human-administered process toward a socio-technical decision system in which algorithms source, rank, screen, assess, and recommend candidates. This integrative critical review examines how AI alters the validity, fairness, legitimacy, and accountability of recruitment decision-making and what forms of human oversight are required when employment opportunities are at stake. The paper synthesizes peer-reviewed empirical studies, established systematic reviews and meta-analytic evidence, foundational research in personnel selection and human–automation interaction, and selected authoritative regulatory sources. It uses a transparent structured literature-identification strategy and critical thematic synthesis across human resource management, organizational psychology, information systems, business ethics, law, and AI governance. The evidence reveals a persistent duality. AI can increase processing capacity, consistency, documentation, reach, and disciplined use of job-relevant information, and under validated conditions may reduce some forms of idiosyncratic human bias. These benefits are conditional rather than inherent. Historical labels, proxy variables, selective outcome data, construct misspecification, opaque vendor systems, automation bias, privacy intrusion, accessibility barriers, and weak contestability can reproduce or amplify disadvantage at scale. Applicant-reaction evidence further indicates that strongly algorithmic procedures can weaken perceived justice, trust, organizational attractiveness, and willingness to pursue employment. The review rejects both technological solutionism and a return to unaided human judgment. It advances a risk-calibrated model of constrained human–AI augmentation in which automation is strongest for administrative and information-processing tasks, whereas consequential exclusion decisions require validated criteria, documented human review, genuine override authority, subgroup monitoring, explanation, accommodation, and reconsideration. Responsible AI recruitment therefore depends less on whether a human is nominally 'in the loop' than on whether the complete decision architecture preserves validity, fairness, accountability, and genuine human agency.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.