Given their accuracy, reliability, and efficiency, algorithms are now a fundamental part of many decision-making processes in both personal and professional domains (Castelo et al., 2019; Prahl & Swol, 2016). They facilitate consumers’ decisions by providing information, suggestions, recommendations, or candidates (Hou & Jung, 2021). However, researchers find that employees are reluctant to use algorithm decision aids, referred to as algorithm aversion (Burton et al., 2020; Dietvorst et al., 2015). The current paper explores the role of self-threat in explaining and overcoming algorithm aversion within a hiring context. We reason that many employees perceive algorithms as threats to their self-concepts because these aids threaten core components of their working selves, such as autonomy, job roles, expertise, status, and job security. As a result, employees respond with a defensive aversion towards these algorithms. Through a randomized experiment, we show that when algorithm decision aids do not pose a threat to the self, employees are more likely to accept these aids and use them. This has implications for the literature on algorithm aversion because we proposed a new way of looking at the underlying cause of this aversion. Our paper also has practical implications as organizations can take steps when introducing algorithm aids to ensure they are not perceived as threats by employees, which can increase their acceptance and usage.
