A new sponsored travel report says only 24 percent of respondents trust AI-generated travel recommendations. Another recent American survey says 65 percent trust AI recommendations at least to some degree, while an Expedia survey found travelers still prefer familiar travel brands when it is time to book.
The numbers do not prove that one survey is wrong. They show that words such as trust, recommendation and booking can produce very different answers, which is exactly why travelers should verify the facts that can break a trip.
- Wunderkind and Skift: 24 percent said they trust AI-generated travel recommendations, and 43 percent felt overwhelmed by AI.
- Qlik and Censuswide: 65 percent of 2,001 U.S. adults planning summer trips said they trusted AI recommendations to some degree.
- Expedia and YouGov: Nearly 70 percent preferred trusted travel brands over AI when booking.
- Why results differ: Samples, wording, market and the stage of the trip-planning process are not the same.
- Traveler rule: Use AI to organize choices, then verify prices, policies, hours, routes and entry rules with the operator.
The new 24 percent figure needs context
The Wunderkind report published with Skift says 24 percent of travelers trust AI-generated recommendations. It also says 43 percent feel overwhelmed by AI, 50 percent are comfortable using it to compare hotels, flights or packages and 40 percent are uncomfortable sharing personal data.
The report’s accessible summary does not provide the full sample size and methodology next to those figures. It is also sponsored content, which does not invalidate the findings but makes transparent methods especially important before the number is treated as a universal verdict.
The behavior pattern is more useful than the single trust percentage. Travelers appear more open to AI during discovery and comparison than during billing, cancellations, disruptions or complicated reservations.
That distinction makes sense. A bad restaurant suggestion is annoying; an incorrect visa rule or missed cancellation deadline can cost thousands of dollars.
Another survey found much higher trust
A Qlik survey conducted by Censuswide asked 2,001 American adults planning summer travel and reported that 65 percent trusted AI recommendations to some degree. The fieldwork ran from June 5 through June 9.
The same survey found that 58 percent used AI for travel planning, 18 percent encountered hallucinations and 15 percent said an AI mistake disrupted or ruined plans. Higher stated trust therefore did not mean error-free use.
Question wording can explain part of the gap. Trusting a recommendation to some degree is a lower bar than giving an unqualified yes to trust, and an American summer-traveler sample is not necessarily the same population as a wider consumer report.
The correct headline is not that trust suddenly rose from 24 to 65 percent. It is that travelers can express moderate confidence while still encountering mistakes serious enough to damage the trip.
Booking remains a separate test
An Expedia and YouGov survey of 5,700 adults in the United States, United Kingdom and India found nearly 70 percent preferred trusted travel brands over AI when booking. Inspiration and transaction confidence are not interchangeable.
A traveler may happily ask an assistant for three beach towns, then book through an airline, hotel or established agency. That is not hypocrisy; it is a reasonable division of labor.
Deep Arrival has tracked the same tension in travelers using AI while still preferring human support. The useful question is not whether someone is an AI believer, but which decisions they are willing to delegate.
| Travel task | AI role and verification source |
|---|---|
| Brainstorm destinations | AI is strong for a starting list. Verify with tourism boards and current local reporting. |
| Compare itinerary shapes | AI is useful for organizing tradeoffs. Verify with operator schedules and maps. |
| Check a visa or entry rule | Use AI only to identify the question. Verify with the government immigration authority. |
| Confirm a theme park hour | AI may summarize, but can be stale. Verify with the official dated calendar. |
| Cancel or change a booking | AI can help read terms. Verify with your confirmation and the provider. |
A simple verification ladder
First, ask the AI for the source behind any time-sensitive claim. If it cannot produce a direct page, treat the statement as an idea to research rather than a fact.
Second, check the date. A real source can still be useless when a parking rate, attraction policy or schedule changed after publication.
Third, confirm that the source controls the fact. An airline is authoritative for its baggage policy, while a blogger may be better at explaining how that policy feels in practice.
Fourth, save the evidence used for a high-cost decision. A screenshot or PDF of the cancellation terms can matter when the live page changes later.
Finally, ask what would happen if the answer were wrong. Low-cost inspiration can tolerate experimentation; passports, medical needs and nonrefundable bookings deserve a human check.
Where AI is already becoming useful
Conversational tools can reduce the blank-page problem. Deep Arrival’s look at Google Maps conversational planning shows how a traveler can move from a vague idea to places positioned on an actual map.
Hotel brands are also adding assistants, including the Hilton planning tools Deep Arrival reviewed. A brand-connected assistant may have better access to that company’s inventory, but it can still favor the brand’s own ecosystem.
The best use is often compression. AI can summarize ten hotel policies into a comparison table, while the traveler opens the final two official pages before paying.
For complicated trips, that final comparison works best inside a repeatable human checklist. Deep Arrival’s conference travel planning framework is one example: let a tool organize the options, but keep the traveler responsible for confirming dates, access rules, transportation and the reservation terms that carry real consequences.
That approach preserves speed without confusing convenience with authority. It also makes errors easier to catch because the traveler verifies a short list instead of the whole internet.
How to read future AI surveys
Look for the sample size, country, dates and who paid for the research. Then find the exact wording of the trust question and whether respondents had recently used the technology.
Check whether the survey measures interest, usage, comfort or completed bookings. Those concepts sound similar in a headline but describe different behavior.
Also look for a denominator. Saying 43 percent of travelers book directly with a brand means something different from 43 percent of travelers who already decided to book.
When methods are missing from the summary, treat the result as directional. It can reveal a topic worth watching without becoming a precise forecast of what every traveler will do.
A three-step verification routine for AI travel advice
First, ask the tool to identify the decision and the authority that controls it. A useful answer should distinguish an airline baggage rule from an airport security rule, or a hotel policy from a government entry requirement.
Second, open the controlling source and check the effective date, location and traveler category. A correct rule for a domestic passenger, resident discount or one park can become wrong when the itinerary changes.
Third, save the confirmation that will matter during the trip. That may be a booking email, official calendar, government page, fare rule or accessibility note with a date and contact path.
This routine keeps AI in the useful part of the workflow. The model can organize questions and summarize tradeoffs while the traveler relies on a source with the authority to honor the answer.
It also makes errors easier to catch because verification is attached to a specific claim. Asking whether an entire itinerary is accurate is vague; checking one terminal, transfer time or cancellation deadline produces a decision.
When a source and an AI answer disagree, use the current source and ask the provider if money or access is at stake. A polished explanation should never outrank the organization that operates the flight, hotel, attraction or border process.
Recheck the answer shortly before travel when schedules, prices or rules can change.
For a complicated trip, keep a short source list beside the itinerary so every important decision can be refreshed without repeating the entire research process.
The conflicting surveys strengthen one practical rule: use AI most freely where mistakes are cheap and verify it most carefully where mistakes are expensive. Destination ideas, packing prompts and comparison tables can save time.
Entry rules, prices, operating hours, accessibility needs and booking terms should come from the authority that controls them.
What to Watch
- Whether future reports publish complete samples and question wording.
- More direct booking functions inside major AI assistants.
- Clearer source links and freshness dates attached to recommendations.
- Travel-provider guarantees for answers generated inside their own tools.
- Independent evidence on how often AI errors cause financial loss.
Trust will probably remain task-specific. Travelers may become more comfortable asking AI where to go while staying cautious about letting it move money or change a reservation.
The product that wins does not need to sound the most human. It needs to show where each claim came from, how current it is and what the traveler should verify before acting.
Do only 24 percent of travelers trust travel AI?
One sponsored report published that figure, but another recent American survey reported 65 percent trusted AI recommendations to some degree. The methods and wording differ.
Why can two surveys disagree so much?
They can use different countries, samples, dates and definitions of trust. Adding the phrase to some degree can also change how people answer.
Is AI safe for trip planning?
It can be useful for ideas and organization. Travelers should verify any fact that affects eligibility, money, timing, health or a nonrefundable booking.
What facts should always be checked?
Verify visas, passports, schedules, prices, cancellation rules, attraction hours, accessibility information and medical guidance with the controlling authority.
What is the best AI travel workflow?
Use AI to create and narrow options, ask it for sources, then open the official pages for the final choices before booking.
