Systemic ethics framework for artificial intelligence in nursing and healthcare

Artificial intelligence (AI) is increasingly integrated into nursing and healthcare practice through clinical decision support, documentation systems, triage tools, predictive monitoring, and other AI-enabled applications. These systems may influence how nurses recognize risk, prioritize care, communicate with patients and families, and exercise professional judgment within interprofessional teams. As AI-supported outputs contribute to clinical decision-making, questions concerning professional accountability, moral agency, documentation responsibility, human oversight, and override authority become increasingly salient. Existing principle-based and governance-oriented approaches provide important normative foundations but may offer limited guidance for addressing how these ethical tensions are negotiated in everyday clinical practice. This study develops a systemic ethics framework for AI in nursing and healthcare through an integrative review. A structured Web of Science and PubMed search (2014-2025) identified 629 records, of which 41 met the criteria for integrative synthesis. Bibliometric mapping was combined with iterative thematic synthesis and conceptual integration to identify recurrent ethical orientations and examine their relationships within patient-care and nursing contexts. Eight review-derived conceptual axes were identified: normative foundations; governance and accountability; trust and transparency; data justice and equity; embedded and lifecycle ethics; education and ethical literacy; generative AI and emerging risks; and patient-centered and relational ethics. These axes informed the development of a multi-level Systemic Ethics Model operating across micro (bedside and professional practice), meso (organizational and health-system governance), and macro (policy, national, and global governance) levels. The model conceptualizes AI ethics as a dynamic, relational, and feedback-sensitive process requiring clear accountability arrangements, meaningful human oversight, continuous ethical monitoring, and institutional support for professional judgment. By integrating relational care values with adaptive governance and lifecycle-oriented oversight, the framework provides a practice-oriented basis for the ethically responsible integration of AI in nursing and healthcare and identifies priorities for future research, education, governance, and implementation.