Abstract

With recent major progress in AI, AI-driven robots are expected to operate outside of their original settings, such as factory floors, and be integrated into our daily lives. In this trajectory, integrating culture, with its broad and vague meaning, into human–robot interaction becomes a concern for the field of social robotics. In this article, we propose a culturally-responsive robot as a concept that encapsulates critical theoretical and experimental studies of culture in robotics without subscribing to the dominant view of the field that translates culture only to nationality. The main contribution of this article is to propose a methodology for implementing culturally-responsive robots, termed “robot-ethnographers.” Drawing from the long history of conventional ethnography, we define a robot-ethnographer as a robot situated within interaction contexts, shaping interactions, and learning culture through observation to adapt its behaviours accordingly. The paper analyses the properties of the robot-ethnographer in line with those of conventional ethnography in social science. Whilst acknowledging important technological limitations, this article concludes that robot-ethnographers can create culturally-responsive human–robot interactions.