From Information Tool to Meaning Mediator: Understanding Consumer Acceptance of GenAI Review Summaries

Abstract

Generative artificial intelligence review summaries (GenAI-RS) are increasingly being embedded in online travel platforms to help consumers process large volumes of reviews, yet the mechanisms explaining their acceptance remain underdeveloped. This study identifies the antecedents of consumers’ acceptance of GenAI-RS and examines how these antecedents shape behavioral intention by using a mixed-method design. Drawing on UTAUT2 and posthumanist and actor–network perspectives, qualitative analysis of semi-structured interviews and online user-generated content identifies two GenAI-RS-specific antecedents: perceived fairness and perceived responsibility uncertainty. A pilot study and the main survey validate the measurement model and test the proposed relationships. Results show that conventional acceptance factors, perceived human-likeness, privacy risk, use-related anxiety, perceived fairness, and responsibility uncertainty significantly shape behavioral intention. This study demonstrates that GenAI-RS acceptance is driven not only by instrumental evaluation but also by normative concerns, advancing acceptance theory and guiding platforms toward fairer, more accountable AI summary design.