63% of travelers would let AI watch for room rate drops. 44% want it planning the trip

Mews has published a second cut of its July traveler survey, and this one is about which AI jobs guests actually want done.

The company asked 3,250 travelers across eight countries what they would use AI for during a hotel stay. Watching for price drops came top at 63%. Real time translation followed at 56%. Planning an itinerary, the use case most travel AI demos are built around, came fourth at 44%. Personalizing the room sat at 41%.

One number explains the ranking: 57% said they would rather stay in control of their trip than hand decisions to AI acting on their behalf. Guests are not asking for an agent that books for them. They want a tool that watches on their behalf and reports back.

What guests will hand over

The data sharing figures follow the same logic. 61% were comfortable with AI using preferences they had stated themselves. 56% would allow past stay history. Online behavior dropped to 43%, and external or social data to 34%. In the US, 63% were happy for AI to use past stay history. In Spain, 72% said transparency and control over their data would make them more willing to use AI services at all.

Consent travels with the source. Data a guest typed in is fair game. Data gathered about them without asking is not.

The markets do not move together

Italy stands out twice. 58% of Italian travelers expect check-in with no front desk, against a 42% average, and 51% are optimistic about AI's role in five years, against 39%. Spain is the most positive market overall. The UK is the quietest: only 30% rate in-stay personalization as important, against a 40% average.

For a group operating across Europe, that is an argument against shipping one guest-facing AI experience to every property.

Josef Starýchfojtů, chief product and technology officer at Mews, said: "Guests have told us where AI helps, and where it doesn't." He described the company's approach as "building for human hospitality: technology handles the mundane, repetitive admin".

Worth knowing before you quote it

This is the same July 2026 survey wave behind the Mews release of 7 October, which found humans preferred over AI at all nine hotel touchpoints tested. The figures here are new. The sample is not. Anyone citing both should say so.

Mews again does not name a research provider, give per-country sample sizes, publish question wording or state a margin of error. And the findings point where vendor surveys usually point. Rate monitoring, translation and personalization are all things Mews sells or plans to.

None of that makes the ranking wrong. It holds across eight markets and it lines up with the touchpoint data from the same wave.

What it means for operators

If you are funding one guest-facing AI feature, this survey says price transparency and language, not trip planning. Those are also the two cheapest to run, because neither has to make a decision for the guest.

If you are already collecting behavioral data to feed a personalization engine, the 43% figure is the one to note. Guests will accept personalization built on what they told you. The further you get from that, the smaller your permission.