Human‐ <scp>AI</scp> Interaction Data Governance in Personal Health Management: Literature and Cases Analysis

ABSTRACT This study investigates data risks in human‐AI interaction within personal health management apps and systems. Drawing on literature, cases, and policies, it analyses data types, governance practices, and potential risks. Using PRISMA (Preferred Reporting Items for Systematic Reviews and Meta‐Analyses) guidelines and thematic analysis, 116 articles, 9 cases, and 4 key policies were examined. Identified data types include physiological indicators, lifestyle habits, personal identity, and social relations. Governance practices and risks are categorised into six dimensions: governance subjects, policies and standards, technologies and facilities, lifecycle measures, data rights protection, and risk management. The study offers governance strategies for human‐centred AI and lays the groundwork for further assessment of risk types and severity.