Is Fairness Enough? The Impact of Expectation Disconfirmation on Job Pursuit Intention in AI-Enabled Recruitment
研究发现，求职者自评资质与AI评分之间的差距（期望不一致）会降低其求职意愿，且程序公平感知部分中介了这一效应，AI信任在某些情况下加剧了负面影响。
Engineering and technology intensive organizations increasingly adopt AI-enabled recruitment tools to improve selection efficiency, but challenges related to algorithmic fairness and trust influence applicant responses. Despite the use of objective and fair algorithms, applicants may still perceive unfairness, but empirical research on this phenomenon remains limited. Based on expectation disconfirmation theory, this study defines the gap between applicants' self-assessed qualifications and AI-generated person-job fit scores as person-job fit expectation disconfirmation and estimates how it affects job pursuit intention through perceptions of procedural justice. To ensure research rigor, we employed a two-stage online scenario-based experiment, implemented attention checks, and conducted exploratory and confirmatory factor analyses to validate construct reliability and validity. Multiple procedural and statistical remedies were also applied to address potential common method bias. Empirical results show that expectation disconfirmation significantly reduces job pursuit intention. Besides, procedural justice partially mediates this relationship, suggesting that weakened fairness perceptions serve as a critical mechanism underlying the negative effect of expectancy disconfirmation. Furthermore, trust in AI exacerbates the negative impact of expectation disconfirmation in some cases. These insights deepen the understanding of how AI-generated evaluations shape individuals' behaviors and perceptions. This study extends the applicability of expectation disconfirmation theory to algorithmic recruitment contexts and offers novel theoretical and practical implications.