Trust-Based Acceptance of AI-Enabled Eco-Friendly Product Recommendations: The Roles of Explainability, Eco-Personalization Fit, and Environmental-Claim Credibility

Abstract

This study investigates how AI-enabled eco-friendly product recommendation features are associated with consumers’ purchase intention in sustainable e-commerce. The research focuses on three antecedents: perceived recommendation explainability, perceived eco-personalization fit, and perceived environmental-claim credibility. Trust in AI-enabled eco-product recommendations is examined as a mediating construct, while purchase intention toward AI-recommended eco-friendly products is treated as the outcome variable. The study is grounded in the Stimulus-Organism-Response framework, supported by Signaling Theory. A structured offline survey was used to collect data from 153 respondents. Python was used for data screening, demographic profiling, descriptive statistics, correlation analysis, and common method bias assessment. Measurement model evaluation, structural model assessment, bootstrapping, and testing of mediation were benchmarked on Smart PLS. The result reveals that eco-personalization fit indeed significantly affects trust and purchase intention, respectively. There is a strong relation between environmental-claim credibility and trust and between trust and purchase intention. Chinese-based recommendation explainability does not significantly relate to trust or purchase intention in the final sample. Environmental-claim credibility also does not show a significant direct relationship with purchase intention, but it has a strong indirect relationship through trust. The results suggest that consumers respond more strongly to recommendation relevance and credible sustainability information than to explanation alone. The study contributes to understanding trust formation in AI-enabled sustainable product recommendation systems and provides practical implications for e-commerce platforms seeking to promote eco-friendly products.