The dark side of AI identity: Investigating when and why AI identity entitles unethical behavior

Abstract

With its rapid development and significant benefits, increasingly more organizations have adopted artificial intelligence (AI) and taken various ways to promote employees’ AI usage and AI-related support behaviors. As a primary way to promote AI usage and AI-related supportive behaviors, building an AI identity is widely recognized as having beneficial effects on employees’ work attitudes and outcomes. Drawing upon role identity theory, we challenge this general conclusion by identifying a potential dark side of AI identity and investigating how and when AI identity promotes unethical behavior. Based on an experiment and a multi-wave field study, we found that AI identity had a positive indirect effect on unethical behavior via psychological entitlement. Furthermore, perceived rarity of AI identity moderated the observed effects—that is, when perceived rarity of AI identity was high, employees with AI identity were more likely to have psychological entitlement, which increased unethical behavior. Taken together, our findings provide new insights into the consequences of AI identity as well as reveal the importance of the rarity of AI identity and psychological entitlement in this process.