Artificial Intelligence (AI) provides unprecedented possibilities to accelerate and improve the hiring process. However, the willingness of recruiters to use–and in turn–profit from AI is crucial. Perceptions of human resource (HR) experts on AI-based hiring has been explored further in an empirical study, which is partly shown in this paper. As the existing body of research is still small, an inductive approach was chosen. We conducted seven semi-structured expert interviews with recruiters from the German manufacturing industry. Based on the findings, we developed a conceptual model with testable hypotheses and practical implications. Noticeably, the recruiters both praised and criticised AI for providing a fact-based assessment of candidates. AI’s increased efficiency was recognised as a major opportunity or even necessity. The most promising approach seems to be a tandem of human and algorithmic decision-making, each offsetting the shortcomings of the other.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Opportunities and Challenges of Artificial Intelligence in Hiring from a Recruiter’s Perspective

  • Maja Dieckmann,
  • Ralf-Christian Härting,
  • Joerg Bueechl,
  • Rosalie Schwörer

摘要

Artificial Intelligence (AI) provides unprecedented possibilities to accelerate and improve the hiring process. However, the willingness of recruiters to use–and in turn–profit from AI is crucial. Perceptions of human resource (HR) experts on AI-based hiring has been explored further in an empirical study, which is partly shown in this paper. As the existing body of research is still small, an inductive approach was chosen. We conducted seven semi-structured expert interviews with recruiters from the German manufacturing industry. Based on the findings, we developed a conceptual model with testable hypotheses and practical implications. Noticeably, the recruiters both praised and criticised AI for providing a fact-based assessment of candidates. AI’s increased efficiency was recognised as a major opportunity or even necessity. The most promising approach seems to be a tandem of human and algorithmic decision-making, each offsetting the shortcomings of the other.