Reading Human Reviews or Relying on AI Summaries? Trust, Diagnosticity, and Information Overload in Online Booking

Şahin M. A., Yurder Y.

Journal of Theoretical and Applied Electronic Commerce Research, cilt.21, sa.11, ss.386-413, 2026 (Hakemli Dergi)

Özet

AI-generated review summaries are increasingly used by electronic commerce platforms to simplify access to consumer feedback. However, consumers’ responses relative to human-generated reviews and the role of information detail remain unclear. Using the Heuristic–Systematic Model (HSM) as an interpretive lens, this study examines trust, perceived diagnosticity, information overload, decision confidence, and booking intention. A 2 × 2 between-subjects experiment with 400 European consumers combined review presentation (human-generated reviews versus an AI-generated review summary) and information detail (low versus high). The data were analyzed using partial least squares structural equation modeling and multigroup analysis. Relative to human-generated reviews, AI-generated review summaries were associated with lower trust and perceived diagnosticity and higher information overload, while their conditional direct effect on booking intention was nonsignificant. Trust and perceived diagnosticity were positively associated with decision confidence, whereas information overload was negatively associated with it. Decision confidence was positively associated with booking intention, with significant indirect pathways from these evaluations through confidence. The AI-generated review summary condition retained a positive conditional direct pathway to decision confidence after controlling for these evaluations. Group differences were concentrated primarily in overload-related pathways under high-detail conditions. The findings highlight opposing pathways within a combined source-format comparison.