AI-Powered Personalization and Online Purchase Intention: A Moderated Mediation Analysis of Trust and Ease of Use

Abstract

This paper explores the role of three mechanisms of AI-based personalization, such as product recommendations, advertisements, and customer support, in influencing online purchase intention among e-shoppers in Pakistan (N = 156). The research is based on an extended Technology Acceptance Model (TAM), in which the notion of trust in AI suggestions is conceptualized as a mediating factor, and the notion of AI-enabling ease of use is formulated as a boundary condition. Through the PROCESS macro analyses, AI-driven product recommendation greatly improved trust (b = 0.44, p < 0.001), and the latter positively influenced purchase intention (indirect effect = 0.20, 95% CI (0.11, 0.33)). The direct impact of AI recommendations on purchase intention was also of significance, and this suggests a partial mediation. The relationship between recommendations and trust had a positive moderating role as ease of use (interaction b = 0.26, p = 0.007), indicating that AI cues are more convincing in an environment where the user feels that the platform is easy to use. The direct effects of AI-powered advertisements and AI-enabled customer assistance on purchase intention were smaller and positive, with no significant gender differences. Overall, the model accounted 34% variation of purchase intention. Considering trust and ease of use as a part of the TAM framework, the study expands TAM to the realm of AI-based personalization and offers a subtle understanding of how smart marketing services should affect consumer decision-making in digital commerce settings.