A critique of the ethical effects of using AI in job recruitment and selection in South Africa.

Abstract

The integration of Artificial Intelligence (AI) into recruitment and selection has transformed human resource management globally, introducing efficiency, precision, and scalability. In South Africa, this transformation occurs within a socio-economic and ethical context shaped by structural inequalities, high unemployment, and constitutional imperatives of fairness and redress. Traditional recruitment processes are often inefficient, with time-to-hire periods exceeding 90 days, and subjective evaluations create a compelling case for AI adoption. This study critically examines the ethical implications of AI-driven recruitment in South Africa through a consequentialist lens, which evaluates actions based on their outcomes rather than their intentions. It explores whether AI enhances welfare, equity, and transparency or perpetuates bias, exclusion, and opacity. Adopting a qualitative, exploratory, desktop-based design, the study uses secondary data from academic, policy, and institutional sources. Applying the DECA method (Describe, Evaluate, Consult, Act) and thematic analysis, it identifies four ethical dimensions: efficiency, fairness, transparency, and candidate dignity. Consequentialist reasoning evaluates AI’s outcomes within South Africa’s legal framework, encompassing the Employment Equity Act (EEA), the Protection of Personal Information Act (POPIA), and the Promotion of Administrative Justice Act (PAJA). Findings reveal that while AI streamlines recruitment and reduces bias, it may also introduce algorithmic discrimination and diminish human oversight. The study concludes that AI’s ethical legitimacy depends on advancing welfare, justice, and dignity. It recommends governance frameworks, capacity-building initiatives, and policies that ensure transparency, accountability, and human control in AI recruitment.