Ex Machina = Ex Manners? Incivility Towards AI and Human Service Agents
Kateryna Reiby, Suzanne van Gils · Academy of Management Proceedings · 2025
People interact with a growing number of digital service agents on daily basis, covering a range of services. These service agents are both human and AI, and communication generally follows patterns of interpersonal interaction. In this paper, we investigate a particular type of human behavior in interactions with AI/human service agents, namely incivility. We study whether people tend to be more uncivil (e.g., by using rude, foul language in communication) towards AI service agents than towards human service agents. We report results of three online between-subject experiments (N=1227) that demonstrate that people have a different civility barometer towards AI service agents as opposed to human service agents. Our findings (study 1) suggest that people evaluate the same verbal expression of incivility less negatively when directed to AI as opposed to humans. We also show (studies 2-3) that people themselves are ruder in response to dissatisfaction with a service when they evaluate service provided by AI service agents (in comparison to human service agents). People also expect higher compensation for poor service if it is provided by AI (study 3). We discuss this differential effect for human and AI service agents in the context of implications for customers and organizations.