AI-Driven Fit Confidence in Fashion E-Commerce: How Personalization, Transparency, and Privacy Concerns Shape Consumer Experience and Purchase Intention

Abstract

Artificial intelligence (AI) is transforming fashion e-commerce through personalized recommendations, virtual try-on, size suggestions, and algorithmic decision support. This study examines AI-driven fit confidence as a fashion-specific mechanism linking AI personalization quality, algorithmic transparency, trust in AI, consumer experience, and purchase intention, while considering privacy concern as a boundary condition. Data from 400 Ecuadorian consumers with prior exposure to AI-enabled recommendation or personalization features in fashion e-commerce were analyzed using partial least squares structural equation modeling. AI personalization quality and trust in AI were positively associated with fit confidence, while algorithmic transparency was associated with trust in AI. Fit confidence was strongly related to consumer experience but showed only a small direct association with purchase intention; its indirect effect through consumer experience was substantially larger, indicating complementary partial mediation. Privacy concern had no significant direct relationship with fit confidence. The interaction between privacy concern and personalization quality was negative, but evidence was mixed and sensitive to the bootstrap confidence-interval method; H7 was not supported under the primary percentile-bootstrap criterion adopted in this study. Predictive analyses showed stronger out-of-sample performance for trust in AI and fit confidence than for purchase intention. The study distinguishes trust in technology from confidence in anticipated product–consumer fit and supports explainable, privacy-sensitive applications in fashion e-commerce.