What did you hear and what did you see? Understanding the transparency of facial recognition and speech recognition systems during human–robot interaction

Abstract

As social robots begin to assume various social roles in society, the demand for understanding how social robots work and communicate grows rapidly. While literature on explainable artificial intelligence suggests that transparency about a social robot’s working mechanism can evoke users’ positive attitudes, transparency may also have negative outcomes. This study investigates the paradoxical effects of the transparency of facial recognition technology and speech recognition technology in human–robot interactions. Based on a lab experiment and combined analyses of users’ quantitative and qualitative responses, this study suggests that the transparency of facial recognition technology in human–robot interaction increases users’ social presence, reduces privacy concerns, and enhances users’ acceptance of robots. However, exposure to both facial and speech recognition technologies revives users’ privacy worries. This study further parses users’ open-ended evaluation of the prospective application of social robots’ tracking technologies and discusses the theoretical, practical, and ethical value of the findings.