Sound familiar? Matching voices boost trust in self-driving cars (news.umich.edu)

A University of Michigan study presented at the Human Factors and Ergonomics Society conference finds that the voice an automated vehicle uses can materially shape public trust. In an online experiment with more than 300 U.S. drivers, participants learned how an AV worked, watched six short driving videos in which the car narrated its actions, and rated their trust after each clip. Researchers distinguished cognitive trust (belief in the car’s competence and reliability) from affective trust (emotional connection) and found that voice matching—especially matching the driver’s gender—boosted both kinds of trust. Age similarity primarily increased affective trust, and when a voice matched gender but not traditional gender-role expectations it mainly increased emotional rather than logical confidence. The findings matter because social acceptance and perceived safety are major barriers to AV adoption. Voice design therefore becomes a low-cost but powerful lever for improving user comfort and comprehension. At the same time, relying on gender-role cues risks reinforcing stereotypes; the authors suggest mitigations such as customizable voices, gender-neutral options, or nonhuman auditory signals. For AV developers and AI/UX designers, the study underscores that sociocultural factors in voice selection have measurable impacts on how people judge automated systems’ competence and trustworthiness, with implications for safety, regulation, and public rollout.