At the heart of the discussions about artificial intelligence (AI) and the robots’ presence in the economy and the job market (van der Zande et al., 2018) lie the anthropomorphized robots, used especially in service industries. Service robots are already being used for elderly care (Kalogianni, 2015), café and restaurant services (Frey & Osborn, 2017) and even for sexual pleasure (Krumins, 2017). Anthropomorphism is defined as imbuing the imagined or real behavior of nonhuman agents with humanlike characteristics, motivations, intentions, and emotions (Epley et al. 2007). In time, the marketplace emotions towards robots turned from excitement and empathy towards anger and fear, due to warnings of several entrepreneurs against uncontrolled AI and increased use of robots for the defense industries (Mattel, 2015). Coinciding with these findings, people are exposed to crash-tests with dummies or defense robots tested against human violence to prove real-time balance (Darrow, 2016). Mende et al. (2017) mention that the rise of service robots is a double-edged sword with increased consumer buzz and engagement on the one hand but uncomfortable feelings on the other. We argue that the level of anthropomorphism, the type of violence towards the anthropomorphic robot and the level of personal characteristics such as power and status play a role in how consumers form and report attitudes towards robots or marketing communication material that depicts such entities. One-hundred and five undergraduates (54 women, mean age = 22.73) took part in our study. Participants first responded to the scales measuring their loneliness, power and status. Participants then reported their mood and moved on to the robotic violence manipulation. Participants were randomly assigned to one of the three photographs depicting a falling robot because of direct human violence, indirect human violence or an accident. Then we measured their evaluation of the material, of the robot, WOMM intentions, their familiarity with the material and their involvement in robotics for control purposes, anthropomorphism level, pre- and post- mood, and demographics. There was a significant effect of the photograph type on attitudes towards the visual (F(2,96) = 3.274; p = 0.043), that is the indirect violence condition evoked more favorable attitudes (M = 4.34; SD = 0.19) than the accident condition (M = 3.65; SD = 0.20, p = 0.013) but not compared to the direct violence condition (M = 3.96; SD = 0.21, p = 0.180). Moreover, there was a marginally significant two-way interaction between power and photograph type on the attitudes (F(1,95) = 3.0021; p = 0.086). The photograph caused a difference in attitudes only if the individual possessed low sense of power (-1SD, 3.97 / 7.00) and not if the viewer had average (M = 4.75 / 7.00) or high sense of power (+1SD, 5.54 / 7.00). People low in power demonstrate variability across contexts (Kraus et al., 2011), which is replicated in our study using anthropomorphized robotic images. Our results are expected to have many implications and pave the way for future work looking at the interaction between the service recipient and the service provider.