Judge a robot by its cover? The impact of delivery purpose labels on willingness to help, perception, and evaluation of robot abuse

Delivery robots not only serve users but also share public spaces with non-users, known as Incidentally Co-present Persons (InCoPs), who might engage in different unplanned interactions. For instance, delivery robots will encounter obstacles on the streets and might require assistance from passersby or might even be attacked by passersby. In two experimental online studies ( total n =329), we investigated how Incidentally Co-present Persons evaluate and respond to delivery robots in public spaces at two levels: passive moral judgment and active behavioral orientation. Study 1 examined whether the delivery purpose of a robot (expressed by a label) influences the moral evaluation of robot abuse. Study 2 investigated how delivery purpose and audio cues (silence, beeping, alarm) shape Existence Acceptance (EA), willingness to help (WtH), expected compensation (EC), and perceived assertiveness, safety, and comfort. The main results of these study are that the delivery purpose did not influence the moral judgment of the attack against the delivery robot. Moreover, though the purpose of the robot and its audio cues did not influence participants’ EA, people were less willing to help a pizza-bot than a pharma-bot. Moreover, individuals with higher EA were also more willing to help and expected less compensation. We further report qualitative results on participants’ reasons for (not) helping a delivery robot.