[AI Virtual Humans Empowering Nursing Education: Opening a New Paradigm for Immersive Teaching]

Nursing education has long relied on either standardized patients or high-fidelity manikins for simulation-based teaching. However, both are constrained by high costs, limited reproducibility of real-world scenarios, and limited opportunities for repeated practice. The utilization of immersive virtual reality pedagogy, AI virtual humans, head-mounted displays, and natural user interfaces offers the potential to create learning environments that are immersive, interactive, and realistic. Related systems support participatory hands-on practice, personalized learning, real-time feedback, and both self-directed and interdisciplinary learning, making it well suited for nursing classrooms and clinical teaching. The process used to develop AI virtual human scenarios using no-code platforms is presented in this article, and the key components and educational outcomes of AI virtual human-based simulation are synthesized. When examined in the context of the four levels of the New World Kirkpatrick Model, existing evidence is concentrated primarily at the "Reaction" and "Learning" levels, with limited assessments of behavior changes or outcomes at the clinical/organizational level conducted. Overall, AI virtual humans may be used to enhance the skills and confidence of nurses in the realms of nurse-patient communication and clinical reasoning/decision-making, and their applications have been gradually expanding to other areas, including suicide prevention and social-emotional learning. In the future, based on appropriate evaluation of platform and equipment costs, more rigorous study designs and multidimensional outcome measures should be incorporated to verify long-term benefits. Also, higher-level clinical implementation effects should be taken into consideration.