Mechanism of public behavioral intention to use generative AI for folk story image co-creation

Traditional folk stories face transmission challenges due to reliance on oral and written preservation, while Generative AI offers new opportunities for innovative dissemination. However, the behavioral mechanisms underlying public engagement in Co-Creating folk story images with Generative AI remain unclear. This study integrates the Technology Acceptance Model and the Theory of Planned Behavior, introducing Evaluation of AI-generated Works (EAW), AI Identity Bias (AIIB), and Self-Efficacy (SE). Using Partial Least Squares Structural Equation Modeling (PLS-SEM) and machine learning on 682 responses, the results show that AIIB exerts a stable linear negative effect on Behavioral Intention (BI) and ranks highest in importance (SHAP = 0.059). EAW demonstrates a nonlinear surge effect and ranks second (SHAP = 0.055), while Subjective Norm, despite its significant positive effect, contributes the least (SHAP = 0.019). Moreover, SE and Perceived Behavioral Control display threshold-dependent positive effects, whereas Perceived Ease of Use and Perceived Usefulness exhibit U-shaped and inverted U-shaped relationships with BI.