Robots are both anthropomorphized and dehumanized when harmed intentionally

This paper investigates the "harm-made mind" phenomenon, where witnessing intentional harm towards agents with ambiguous minds (like robots) leads to increased mind perception in those agents. Two experiments replicated previous findings, showing participants attributed higher pain capacity to harmed robots. The study also explored whether robots' emotion-simulation abilities affected this effect and whether harming a robot impacted the perceived prosociality of the harmer. Results showed a positive indirect effect of harm on mind perception via perceived pain capacity, but a negative direct effect, suggesting both anthropomorphism and dehumanization. Harming a robot was perceived as less prosocial.

Marieke S. Wieringa, Barbara C. N. Müller, Gijsbert Bijlstra, Tibor Bosse