University students' learning intention towards generative art in higher education: A SEM–ANN study of human–AI co-creation factors

The pervasive implementation of generative AI technology in art and design has made exploring students' learning intentions towards generative art a pivotal research topic in higher education. This study aims to explore the factors influencing university students' intention to learn generative art, providing a theoretical basis for teaching reform. A combination of Partial Least Squares Structural Equation Modelling (PLS-SEM) and Artificial Neural Network (ANN) hybrid analysis was employed to analyse data from 556 valid questionnaires collected from 58 Chinese universities with art and design programs. A multi-dimensional factor prediction model was constructed, incorporating Cognitive Expectancy (CE), Learning Expectancy (LE), Social Context Influence (SCI), Affective Motivation (AM), Human-AI Co-creation Experience (HACE), and Ethical Responsibility Awareness (ERA). The findings indicate that AM and HACE exert the most significant influence on students' learning intention, with normalized relative importance values of 100 % and 91.6 %, respectively. This is followed by ERA (81.5 %), LE (77.7 %), CE (77.6 %), and SCI (76.9 %). The model's R² value is 66.3 %, and its Q² value is 0.657, indicating strong predictive power. The findings provide theoretical guidance and practical insights for generative art education, suggesting that teaching should focus on affective motivation and human-AI co-creation experience while integrating ethical responsibility education. Establishing a clear learning path will enhance the appeal of generative art, promote students' learning intention and creative innovation, and provide theoretical support and empirical reference for the development and expansion of generative art courses in higher education.