The aim of this paper is to present an approach to the human-centered design of AI-driven robots for healthcare through a case study of AIREC (AI-driven Robots for Embrace and Care), a humanoid robot developed under the JST Moonshot Program. To deepen understanding of the social impact of AI-driven robots, I introduce the “complexity model of communication,” with four types of human–AI interactions. The paper also highlights both similarities and differences in perceptions of AI and robotics in healthcare, drawing on cross-cultural research that includes survey data from over 6,000 young people in the United States, United Kingdom, and Japan, as well as in-depth interviews with more than 300 young people across seven countries (the US, UK, Japan, China, Chile, Spain, and Singapore). Finally, I present the Smart Robot Ethics Basic Policy and a corresponding checklist designed to guide engineers in developing human-centered, ethically grounded AI-driven robots for healthcare.