Abstract
Artificial intelligence (AI)-powered personalization algorithms have transformed contemporary e-commerce by curating hyper-relevant product recommendations tailored to individual behavioral footprints. However, the precise cognitive mechanisms linking visual attention to behavioral intentions in algorithmic retail environments remain incompletely understood. This study investigates the influence of AI-driven visual curation on consumer purchase intent within online fashion retail using a converging-methods approach combining laboratory-based eye-tracking (N = 72) and an extended post-task behavioral survey (N = 310). Utilizing a controlled between-subjects experimental design, participants interacted with either an AI-personalized or a non-personalized mock fashion platform. Objective oculomotor metrics—including Total Fixation Duration (TFD) and Time to First Fixation (TTFF) on personalized Areas of Interest (AOIs)—were integrated with partial least squares structural equation modeling (PLS-SEM) of psychological constructs. The results demonstrate that AI-curated interfaces significantly reduce visual search friction, as indicated by faster TTFF and concentrated TFD on recommended apparel items. Furthermore, visual processing fluency positively predicts perceived algorithmic relevance and shopping enjoyment, which sequentially drive purchase intention. Importantly, consumer brand trust was identified as a critical mediator, whereas perceived privacy risk exerted a boundary condition dampening behavioral conversion. These findings bridge human-computer interaction and consumer cyberpsychology, offering actionable insights for the ethical design of visually optimized, AI-enhanced digital storefronts.
