Abstract: Objective: To construct a multi-modal Artificial Intelligence Standardized Patient (AISP) system for nursing education and to inform the intelligent transformation of nursing pedagogy. Methods: A multi-modal AISP system was developed based on the Nursing Process Theory and Constructivist Learning Theory, integrating natural language processing, speech recognition, and visual generation technologies. A total of 118 nursing undergraduate interns at a tertiary grade A hospital in Guangxi were selected as the research subjects to evaluate the system's effectiveness in enhancing clinical interviewing skills, medical record writing proficiency, and learning satisfaction. Results: Following AISP-based training, students' total clinical interviewing score significantly increased from 163.31±11.57 to 190.21±6.61 (P<0.001). Their medical record writing score also showed marked improvement, rising from 79.25±4.33 to 90.74±2.67 (P<0.001). Furthermore, student satisfaction significantly increased across all evaluated domains (P<0.05). Conclusion: The developed multi-modal AISP system effectively improves nursing students' core clinical competencies and learning satisfaction. It presents a feasible and promising paradigm for driving intelligent innovation in nursing education.

Key words: Artificial Intelligence; standardized patient; multi-modal; nursing interns; teaching application