When appearance meets personality: robot age and individual differences in the evaluation of cheating robots

As robots increasingly engage in autonomous and socially meaningful actions, understanding how humans evaluate morally ambiguous robot behaviour is a central challenge in Human–Robot Interaction (HRI). This study investigates how robot age cues and individual differences influence the evaluation of deceptive behaviour displayed by robots in immersive virtual reality. Drawing on theories of mind perception and moral agency, we examine whether Dark Triad personality traits and lie acceptability shape tolerance toward robot cheating and interact with robot age (adult-like vs. child-like) in predicting perceived agency, experience, moral responsibility, and trust. Participants interacted with a humanoid robot that engaged in systematic cheating during a competitive rock–paper–scissors task in virtual reality. Results (𝑁 = 56) show no significant main effect of robot age on moral or cognitive evaluations, suggesting that explicit cheating behaviour may override appearance-based cues. However, significant interactions emerged between robot age and individual differences. Exploratory analyses suggested that narcissism was positively associated with perceived agency and experience, particularly in the adult-like robot condition. Similarly, lie acceptability predicted higher perceived experience and lower dependency exclusively in the adult-like condition. These findings suggest that the evaluation of robot misconduct depends on the interaction between design features and observer dispositions rather than on robot characteristics alone. Adult-like robots appear to amplify individual differences, whereas child-like robots may attenuate them. The study contributes to HRI research by highlighting the importance of considering both user traits and design cues in morally sensitive interactions, with implications for the development of socially acceptable artificial agents taking into consideration ethical concerns regarding the strategic use of dark patterns in the field of social robotics. Ultimately, these findings advocate for HRI design standards that prioritize behavioral transparency and ethical accountability.