Travelbee and relational AI: A triadic model for nursing.

BACKGROUND: Artificial intelligence (AI) in health care is typically evaluated on efficiency, predictive accuracy, and clinical decision support. Far less attention has been paid to how it reconfigures the conditions under which human beings meet one another in situations of illness, suffering, and care. Nursing, grounded in relational and existential understandings of practice, offers a distinct vantage point on this shift. PURPOSE: To examine relational AI, understood as conversational systems designed to establish and sustain socio-emotional bonds with users, through Joyce Travelbee's Human-to-Human Relationship Model and to explore how relational AI may reshape relational processes in nursing care. METHODS: This theoretical paper draws on conceptual analysis supplemented by an integrative literature synthesis. Two complementary searches addressing Travelbee scholarship and relational AI were conducted in MEDLINE, CINAHL, Scopus, Web of Science, PsycINFO, and IEEE Xplore. The final corpus comprised 28 sources, cross-read through Travelbee's five relational phases using a hermeneutic strategy. DISCUSSION: We argue that relational AI introduces a third presence into what Travelbee conceived as a dyadic encounter, reshaping trajectories of meaning, hope, and recognition. We propose the Triadic Human-to-Human Relationship Model (T-HHRM), articulated through six propositions and three modes of mediation, with phase-differentiated guidance for practice. T-HHRM is advanced as a proposed conceptual framework requiring empirical testing, not as an established or validated model. CONCLUSION: T-HHRM offers a conceptual basis for examining AI-mediated relational care and frames implications for AI design and governance, nurses' ethical responsibility, and empirical inquiry into meaning, trust, and hope.