Gen AI or Human? Impact of the Agent Type on Consumers’ Adoption of Alternative Customized Itinerary

Consumers sometimes encounter recommendations that deviate from their initial preferences. In such expectation-discrepancy situations, consumers often infer the underlying motives of the recommending agent. This research examines how agent type, generative AI versus human agents, shapes consumers’ adoption of alternative customized itineraries in customized travel services. Across five online experiments, we find that AI-generated alternatives lead to higher adoption than those suggested by human agents. This effect is mediated by ulterior motive attribution. Consumer knowledge strengthened this effect, whereas providing a fact-based explanation attenuated the difference between AI and human agents. This research contributes to the literature by clarifying why and when generative AI differs from human agents and by identifying motive-based attribution as a key mechanism. The findings also suggest that tourism firms can strategically deploy AI agents in expectation-discrepant service recommendation contexts and tailor recommendation strategies based on travelers’ knowledge levels and the availability of fact-based explanations.