Research on influencing factors of college students' dependence on AI painting tools: A hybrid method based on SEM-ANN.

With the recent advances in technology, AI painting tools have become a mainstream part of digital creation. However, limited research has examined students' dependence on AI-powered drawing tools. Therefore, this study uses college students as a case study to analyze the key factors influencing dependence on AI painting tools. Drawing on the I-PACE model, we constructed a dependence model for AI painting tools. The model incorporates creative pressure, self-efficacy, positive experience, cognitive inertia, and information processing dependence. The study employs a mixed-methods analytical approach combining structural equation modeling (SEM) and artificial neural networks (ANN), and is based on 320 valid questionnaires from university students in China. The SEM results showed positive correlations between creative pressure, positive experience, cognitive inertia, information processing dependence, and AI painting tool dependence. No significant correlation was found between self-efficacy and dependence on AI painting tools. According to the ANN findings, information-processing dependence was the most important predictor of dependence on AI painting tools (normalized importance = 100%), followed by cognitive inertia (80.6%), creative pressure (33.6%), and positive experience (20.1%). These empirical findings provide new insights and evidence into college students' reliance on AI painting tools. Based on this, it is recommended to address students' AI dependence, promote a shift from tool dependence to rational application, and ultimately achieve human-machine collaborative development.