A Rapid Systematic Review of AI-Generated and Human-Generated Real-Time Feedback in Nursing Education

To explore the effects of AI-generated and human-generated real-time feedback on learning outcomes and experiences among undergraduate nursing students.

Timely, high-quality feedback is crucial in nursing education to support clinical reasoning, skill development, and student engagement. While human-generated feedback remains central to pedagogy, emerging AI offers real-time feedback that complements traditional approaches. Understanding how both interventions influence student learning is critical to informing future curriculum and clinical practice.

We conducted a rapid systematic review of quantitative, qualitative, and mixed-methods studies across major databases from 2018 to 2025. Eligible studies involved undergraduate nursing students receiving real-time feedback generated from AI or humans. The Sample, Phenomenon of Interest, Design, Evaluation, and Research Type framework guided inclusion, and the Mixed Methods Appraisal Tool assessed methodological quality.

Nine studies were included. AI-generated tools supported timely, consistent, and individualised learning, contributing to student engagement, confidence, and skill development. Educator-mediated feedback, including technology-assisted and conventional human feedback approaches, supported student satisfaction, reflection, and emotional support. One study directly comparing AI-generated and human-generated feedback found that AI-generated feedback offered greater consistency and objectivity, whereas human-generated feedback provided contextual sensitivity and interpersonal connection; however, this comparison should not be generalised across the wider evidence base.

AI offers opportunities to address scalability and access challenges in feedback delivery but does not fully replicate the relational aspects of educator feedback.

AI-generated and human-generated feedback demonstrate distinct yet complementary strengths. A hybrid feedback model is proposed as a plausible approach that may optimise educational outcomes in nursing; however, the included studies have not directly evaluated it, and future empirical testing is needed before it can be recommended as the most effective approach.

Combining AI-generated and human-generated feedback demonstrates complementary strengths, and integrating them may optimise learning outcomes.

Investment in digital education and faculty development is needed to support scalable feedback strategies globally.

This systematic review adhered to relevant EQUATOR guidelines and followed the PRISMA guidelines (see Appendix 3).