The MME uses trolley-type dilemmas that force people to choose between killing one person (or set of people) versus killing another person (or set of people). Because this paradigm assumes inequality (for example, should we program AVs to kill men or women?), it has difficulties revealing whether people prefer equality (for example, should we program AVs to ignore gender?). The forced inequality condition was a simplified replication of the MME, testing whether participants thought autonomous vehicles should (1) kill group A (for example, elderly people) to save group B (for example, children) or (2) kill group B to save group A. As in the MME, we examined both personal features (for example, kill men versus women) and structural features (for example, kill many people versus few people) in driving situations. In study 3, participants (N = 993 US participants from an online panel) chose which of two autonomous vehicles should be allowed on the road: one that makes ethical decisions on the basis of the structural features revealed by the MME (for example, saving more people versus fewer, killing by inaction versus action), and another on the basis of both structural and personal features (for example, saving people based on age, gender, and status). ...we recognize that people often do discriminate on the basis of personal features, as sexism, classism, racism and ageism all illustrate.