The Effect of AI-Based Recommendations on Embedded Finance Use in E-Commerce through Trust Mediation among Generation X in Jakarta

Artificial intelligence increasingly shapes how e-commerce platforms recommend integrated financial services, yet evidence concerning Generation X remains limited. This methodological study examines the relationships among perceived AI-based recommendation quality, trust, and embedded finance use, while also distinguishing recommendation quality from exposure frequency. A structured synthetic dataset of 300 observations representing Generation X e-commerce users in Jakarta was analysed using partial least squares structural equation modelling with 10,000 bootstrap subsamples. AI-based recommendations were positively related to embedded finance use (β = 0.514) and trust (β = 0.444), while trust was positively related to use (β = 0.240); all paths had p < 0.001. The indirect effect through trust was significant (β = 0.107), indicating complementary partial mediation, and the model explained 43.1% of the variance in use. A contextual cross-tabulation showed that recommendation exposure frequency was not associated with purchase response (p = 0.818). The study's novelty lies in showing that, for digitally experienced Generation X profiles, perceived relevance and accuracy matter more than repeated exposure, while trust strengthens but does not dominate the relationship. Empirical validation with linked respondent-level data remains necessary.