Ethical, Legal, and Regulatory Implications of Artificial Intelligence in Healthcare: A Narrative Literature Review

Introduction: Artificial intelligence (AI) is rapidly spreading across healthcare settings, offering tools that support diagnosis, clinical decision-making, and personalised care. Its adoption, however, raises relevant ethical, legal, and regulatory questions related to privacy, data security, algorithmic transparency, bias, and professional accountability, which are particularly critical for nursing practice. Objective: To analyse, through a narrative review of the literature, the main ethical, legal, and regulatory implications of AI in healthcare, with particular attention to governance, privacy, transparency, accountability, equity, and the role of healthcare professionals. Methods: A narrative review, reported in accordance with the Scale for the Assessment of Narrative Review Articles (SANRA), was conducted using a structured search of PubMed, Cochrane Library, Scopus, and CINAHL for literature published in Italian or English between January 2020 and June 2026. The review question was structured using the Population-Concept-Context (PCC) framework. Two reviewers independently screened records and assessed full texts, and a third reviewer resolved disagreements. Study selection is reported with a PRISMA 2020 flow diagram. Results: Of 7906 records identified, 11 studies were included (9 from databases and 2 from reference lists): 3 systematic reviews and 8 narrative reviews, concept analyses, or theoretical papers. The evidence base was therefore predominantly review-level and conceptual. Across the literature, the strongest convergence concerned the need for data protection, human oversight, bias mitigation, and explicit accountability. Unresolved questions concerned the level of explainability required in practice, the allocation of responsibility among developers, organisations, and professionals, and the governance of generative AI. Conclusion: The synthesis suggests that responsible AI implementation depends less on adding isolated ethical safeguards than on clearly assigning responsibilities across the AI life cycle. For nursing, the literature points to the need for AI literacy, the authority to question or override outputs, organisational escalation pathways, and the preservation of professional judgement and patient autonomy. Because the evidence base is predominantly conceptual, these conclusions are interpretive and require empirical testing.