AI-driven product recommendation systems have become an important feature in digital commerce, helping users discover products, reduce information overload, and make online shopping decisions more efficiently. However, technically advanced recommendation systems are not always trusted, understood, or reused by users, especially when recommendations are perceived as repetitive, unclear, or difficult to control. This study aims to analyze the factors influencing users’ reuse intention for information generated by AI-driven product recommendation systems in digital shopping platforms. The research was conducted by collecting data from users of digital shopping platforms through an online questionnaire using Google Forms. This study developed a structural model involving eight constructs: Perceived Algorithmic Accuracy, Perceived Recommendation Diversity, AI-Driven Recommendation Transparency, Perceived Usefulness, Perceived User Control, Trust in AI-Driven Recommendations, Recommendation Satisfaction, and Reuse Intention for Information. A total of eight hypotheses were formulated and tested using Partial Least Squares Structural Equation Modeling (PLS-SEM) through SmartPLS 4. The results show that all proposed hypotheses are supported, indicating that each relationship among the constructs is statistically significant. Trust in AI-Driven Recommendations emerged as the strongest predictor of Recommendation Satisfaction, while AI-Driven Recommendation Transparency strongly influenced Perceived User Control. The findings suggest that successful AI recommendation systems should not only focus on algorithmic accuracy, but also provide diverse recommendations, transparent explanation cues, and user-control mechanisms to strengthen trust, satisfaction, and continued reuse of recommendation-related information.
Halim et al. (Mon,) studied this question.