Unjust algorithmic recommendations can lead decision makers to discriminatory choices, risking harm to individuals or groups. This study addresses this concerning phenomenon and examines its implications. In an experimental study involving 122 managers, we found that algorithmic injustice causes discriminatory decisions without heightened guilt perception. Additionally, trust in data analytics outcomes moderates the impact of algorithmic injustice on discrimination and marginally influences the impact of discriminatory decision making on guilt perception. However, displacement of responsibility has no moderating effect on either relationship. These findings highlight the potential negative consequences of algorithmic decision making, emphasizing the need for caution and awareness.

不公正的算法建议可能会导致决策者做出歧视性选择，从而冒着对个人或群体造成伤害的风险。本研究解决了这一令人担忧的现象并探讨了其影响。在一项涉及 122 名管理者的实验研究中，我们发现算法不公正会导致歧视性决策，但不会增加负罪感。此外，对数据分析结果的信任会缓和算法不公正对歧视的影响，并略微影响歧视性决策对内疚感的影响。然而，责任的转移对这两种关系都没有调节作用。这些发现强调了算法决策的潜在负面后果，强调了谨慎和认识的必要性。