When Health Becomes Work: Ethical, Legal, and Social Issues of AI-Powered Digital Health Services for Employees

抄録

This concept paper introduces, for an international readership, an exploratory identification of the ethical, legal, and social issues (ELSI) associated with the implementation of artificial intelligence (AI)-powered digital health services for employees, first reported in Japanese, and extends it by situating the issues in international regulatory and scholarly contexts and by deriving research questions. Digital health services are proliferating, and AI is increasingly embedded in their data acquisition, analysis, and interactive feedback. Employers are now beginning to provide such services to their employees; because employment relationships involve power asymmetries, workplace implementation raises ELSI distinct from those of privately chosen use. Drawing on three service types piloted in Japan (prediction of future health status based on health checkup results, health support using 3D body avatars, and AI posture assessment with scoring and ranking), a hypothetical workplace scenario was constructed, and issues were extracted through a four-step process: review of published AI-safety evaluation guides, generative-AI-assisted exploration, structured discussion at an ELSI research center, and cross-checking against the Assessment List for Trustworthy Artificial Intelligence (ALTAI). Thirty-one issues were identified and organized into seven categories. Three cross-cutting themes are discussed: extension of the 'right to disconnect,' uniform demands for 'health' under power asymmetry, and employment-status disparities. The resulting list offers a shared starting point for employers, employees, and service providers.