As the German New Wave band Ideal once sang, “Das ist gefährlich, lebensgefährlich, zu viel Gefühl” (“It’s dangerous, potentially fatal, feeling too much”). At least that seems to be how scientists at Radboud University Nijmegen, Netherlands, and in Munich see robots whose design mimics humans. The researchers conducted a trial to find out whether subjects would put an individual in mortal danger to save a group of injured people. This individual was sometimes a human, sometimes a human-like robot, and sometimes obviously just a machine. The result was astonishing : The more intensely the robot was perceived as a living being with feelings, the harder it was for the test subjects to sacrifice it – despite the fact that, within the structure of the experiment, that meant risking multiple human lives. This indicates that robots have been granted some kind of moral status, says developmental psychologist Markus Paulus of the University of Munich. Depending on the area of application for the robot, in fact, it might even be counterproductive to design them to look as similar to humans as possible.

The fact that robots can spark emotions and empathy from humans had previously been demonstrated in a study at the University of Duisburg-Essen. In that experiment, many subjects hesitated to switch off a human-like machine when it implored them not to. Accordingly, there are consequences if machines are equipped with human patterns of behavior, according to the scientists: “You have to wonder whether it is ethically desirable.”