Bias in AI Hiring Tools

Impacted Groups, Legal Risks, Historical Foundations, and Next Steps

Abstract

This paper investigates the role and influence of artificial intelligence (AI) in applicant tracking systems (ATS) on marginalized groups within the course of the job recruitment process. Although AI-powered ATS may ensure efficiency in recruitment through automated resume screenings and interview analysis, it extends the circle of historic bias, which affects immigrants, persons with disabilities, women, and those with non-Anglo names. These systems tend to screen out qualified candidates for non-standard language, gaps in employment, or characteristics irrelevant to job performance. These practices only further perpetuate economic disparities and psychological harm within already marginalized communities. Notable cases involving such firms as Amazon and Workday demonstrate the legal consequences connected with these discriminatory practices, showcasing the need for organizations to track AI bias to avoid legal liabilities. We call for a review of hiring practices and an inclusive redesign of AI systems that holistically evaluates diverse candidate experiences. It necessitates a paradigm shift in AI ethics, inclusive of active participation of marginalized groups in development, to foster equitable hiring practices through recruitment technologies.