Performance and moral trust in care robotics: ethical foundations for human–robot relationships

Abstract As global populations age and birth rates decline, many societies, especially in developed countries, face critical labor shortages in the caregiving sector. This demographic imbalance is straining existing care infrastructures and raising urgent questions about how to support an increasing number of elderly individuals who can no longer live independently. Socially interactive AI systems, particularly care robots, have emerged as a promising technological response. However, their successful integration into real-world care settings hinges on more than technical capability; it requires appropriately calibrated reliance on robotic performance and morally trustworthy sociotechnical care arrangements. This paper investigates the ethical and conceptual foundations of trust in human–robot interactions, with a focus on the caregiving context. Building on the distinction between performance trust and moral trust, we argue that its application to care robotics reveals an agency-trust mismatch: care robots are not moral agents, yet care contexts generate morally significant expectations concerning dignity, autonomy, privacy, non-deception, responsibility, and well-being. We therefore reconceptualize moral trust in care robotics as sociotechnical moral trust, directed not toward the robot alone but toward the broader care arrangement in which robotic behavior is designed, deployed, monitored, governed, and corrected by human and institutional actors. On this basis, we distinguish robot-level performance trust from system-level moral trustworthiness and derive ethical design and governance principles for trustworthy care-robotic systems.