Abstract
OBJECTIVES: This study aims to systematically investigate the application of Generative Artificial Intelligence (GAI) in nursing practice within China. It seeks to map current usage patterns, identify perceived benefits and implementation challenges, and uncover the functional needs of nursing staff regarding GAI. Additionally, the research will assess the real-world performance and adoption of emerging local GAI platforms. The findings are expected to provide foundational evidence to guide the scientifically sound and contextually appropriate development of GAI in nursing. METHODS: The convenience sampling method was used to select nurse interns, staff nurses, and nurse managers from 20 provinces between April and June 2025 as survey respondents. We designed the questionnaire through a literature review as well as evidence extraction. A panel of experts assessed the content validity of the questionnaire. RESULTS: According to a survey of 181 nurses, GAI has achieved a high adoption rate (92.81%), which was not significantly associated with any demographic factors (all p > 0.05). Further analysis, however, revealed that seniority was a significant predictor of usage frequency (B = 0.507, p = 0.001). In terms of platform preference, locally developed GAI platforms, such as DeepSeek and Doubao, dominated the landscape, collectively accounting for over 60% of total usage. The most prevalent application was the generation of health education materials (61.30%). A key finding to emerge from the data is the role-specific utility of GAI: it reduced administrative workload for nurse managers, enhanced workflow efficiency for clinical nurses, and fostered innovation among nursing interns. Despite this promising adoption, significant challenges remain, primarily concerns regarding the professional accuracy of AI-generated content (40.96%) and its potential impact on clinical autonomy. CONCLUSION: This study establishes that GAI adoption among Chinese nurses is widespread but nuanced, marked by a preference for local platforms and role-specific benefits. The primary barriers to integration-concerns about content accuracy and clinical autonomy-underscore the need for clinically validated and trustworthy systems. These findings advocate for targeted training and ethically aware design to fully realize GAI's potential as a collaborative tool in nursing.