Mindless Robots get Bullied

Abstract

Humans recognise and respond to robots as social agents, to such
\nextent that they occasionally attempt to bully a robot. The current
\npaper investigates whether aggressive behaviour directed towards
\nrobots is influenced by the same social processes that guide human
\nbullying behaviour. More specifically, it measured the effects of
\ndehumanisation primes and anthropomorphic qualities of the robot
\non participants’ verbal abuse of a virtual robotic agents. Contrary
\nto previous findings in human-human interaction, priming participants
\nwith power did not result in less mind attribution. However,
\nevidence for dehumanisation was still found, as the less mind participants
\nattributed to the robot, the more aggressive responses they
\ngave. In the main study this effect was moderated by the manipulations
\nof power and robot anthropomorphism; the low anthropomorphic
\nrobot in the power prime condition endured significantly
\nless abuse, and mind attribution remained a significant predictor for
\nverbal aggression in all conditions save the low anthropomorphic
\nrobot with no prime. It is concluded that dehumanisation occurs
\nin human-robot interaction and that like in human-human interaction,
\nit is linked to aggressive behaviour. Moreover, it is argued
\nthat this dehumanisation is different from anthropomorphism as
\nwell as human-human dehumanisation, since anthropomorphism
\nitself did not predict aggressive behaviour and dehumanisation of
\nrobots was not influenced by primes that have been established in
\nhuman-human dehumanisation research.