An Examination of AI in Travel Planning Across Traveler Spending Segments - Cornell Research

Here's an executive summary and key takeaways for the Cornell Center for Hospitality Research report.

Executive Summary

This Cornell Center for Hospitality Research report, authored by Young Jang and Christopher Anderson, examines how U.S. travelers use generative AI in travel planning across four spending segments—Budget, Premium, Aspirational, and Luxury. Drawing on a survey of 1,029 active travelers segmented by their actual nightly accommodation spend rather than demographics, the study finds that AI adoption is far from uniform. Travelers most readily use AI for "discovery" tasks such as identifying activities, attractions, and answers to factual questions, but they remain hesitant to hand over more subjective, complex, or high-stakes decisions. The central thesis is that AI's value proposition and the conditions for trust differ meaningfully by segment, undermining the assumption that a single, one-size-fits-all AI tool will serve the entire market.

Across all four groups, the dominant barrier to broader adoption is consistent: concern about the accuracy of AI-generated information, cited by more than 60% of respondents, followed by worries about transparency and overly generic recommendations. Yet the motivations driving adoption diverge sharply. Budget travelers treat AI as a value-verification engine, Premium travelers use it as a discovery and optimization tool, Aspirational travelers lean on it as a curation assistant for personalized recommendations (contingent on data privacy), and Luxury travelers want AI for fast factual research while reserving final trip orchestration for human advisors. The authors conclude that the future of AI in travel lies in enhancing—not replacing—existing planning behaviors, and that hospitality firms can convert AI from a "risky novelty" into a "trusted conversion partner" by aligning tool design with each segment's distinct expectations.

8 Key Takeaways

1. AI adoption is segment-specific, not universal. A traveler's spending tier—shaped by financial capacity, behavior, and psychology—strongly predicts how and why they use AI, so a single approach won't satisfy the whole market.

2. Discovery is AI's sweet spot. Travelers across all segments are most comfortable using AI to find things to do, surface attractions, and retrieve factual travel information; trust drops sharply for subjective or complex recommendations.

3. Accuracy concerns are the universal barrier. More than 60% of respondents cite worries about factual correctness as the top reason for hesitation, with lack of transparency and overly generic recommendations (each cited by 40%+) close behind.

4. AI still ranks behind traditional tools. When asked for their top three planning tools, respondents ranked Search first, followed by Reviews, Official Hotel Websites, then AI Chatbots/Assistants (4th), and OTAs—signaling AI is established but not yet dominant.

5. Budget travelers use AI to verify value. With over 60% rating price "extremely important," this segment (36.3% of the sample) wants value-first interfaces, real-time price comparison, deal alerts, and transparent source citations.

6. Premium and Aspirational travelers prioritize quality and curation. Premium travelers (the largest group at 39.9%) use AI as a discovery/optimization engine to vet options, while Aspirational travelers rely on it for personalized hotel curation—but only with assurances of data privacy and security.

7. Luxury travelers see AI as an enabler, not a replacement. This segment values AI for rapid fact-finding but resists using it for complex logistical coordination, preferring human advisors for the final, bespoke orchestration of a trip.

8. The opportunity is augmentation, not replacement. AI should enhance existing planning behavior; firms that tailor design to each segment—efficiency for Budget/Premium, discovery and personalization for Aspirational, human-augmenting support for Luxury—can build genuine trust and conversion.

Read the full article here.