Artificial intelligence (AI) and machine learning (ML) are increasingly integrated into healthcare systems, with significant implications for nursing practice. However, evidence regarding effectiveness, implementation challenges, and impact on nursing outcomes remains fragmented. This narrative review, using PubMed, CINAHL, Scopus, and IEEE Xplore (from Jan 2020-Dec 2025), examines AI/ML applications in nursing practice, analyses implementation challenges, evaluates clinical effectiveness evidence, and identifies research gaps. Search terms combined 'artificial intelligence,' 'machine learning,' 'nursing,' and 'clinical decision support.' Inclusion: peer-reviewed empirical studies examining AI/ML in nursing. Exclusion: theoretical papers, nonnursing studies. Analysis employed GRADE criteria for evidence quality assessment; 68 studies demonstrated AI/ML applications across predictive analytics, clinical decision support, intelligent monitoring, documentation, and education. Critical limitations included small samples, short term follow-up, limited RCTs, and algorithmic bias affecting underrepresented populations. Evidence quality was predominantly moderate. While AI/ML demonstrate technical capability, evidence for sustained clinical effectiveness remains limited by methodological weaknesses, implementation challenges, and insufficient examination of nursing-sensitive outcomes. Critical engagement, rigorous evaluation, and attention to equity are essential.