Your Guests Are Two Years Ahead of You on AI — And They Still Want a Human

There's a quiet role reversal happening in our industry, and this week the research finally said it out loud. For most of the last two decades, hotels led and guests followed — we built the booking engine, they learned to use it; we added the app, they downloaded it. On artificial intelligence, that order has flipped. New research making the rounds this week carries a blunt headline: travelers have surpassed the travel industry by roughly two years in AI adoption. The guest arriving at your desk has, in many cases, already used AI more fluently to plan their trip than your property uses it to run. That's not a threat. But it is a fact worth staring at, because it quietly rewrites what a guest now expects of you.

The numbers underneath the headline are striking. In Amadeus's "Tailored Horizons" study, published this month, 74% of U.S. travelers said they'd already used AI while planning a trip — for hotels, routes, transportation — and 85% agreed the tools save them time and give more personalized suggestions. Phocuswright put generative-AI trip planning at around 40% of U.S. travelers back in 2025. A separate outlook survey found a third of travelers plan to lean on tools like ChatGPT, Gemini or Claude for their 2026 trips, using them mostly to get recommendations (75%), build itineraries (70%) and compare options (55%). Whatever the exact figure, the behavior is now mainstream: by the time a guest talks to you, they've often already had a long conversation about your destination with a machine.

The part everyone misses: they don't actually want to hand over the keys

Here's where the story gets interesting, and where a lot of the industry hype falls apart. The same guests who plan fluently with AI are remarkably unwilling to let it act for them without supervision. In the Amadeus data, only 20% would allow an AI to spend money on their behalf without approval; 43% would permit it only with conditions and restrictions, and 29% wouldn't allow autonomous spending at all. Only 13% of the AI users in that outlook survey had used it to actually book. Skift's read on leisure travelers is starker still: just about 2% currently grant an AI full booking authority. And when something goes wrong, the preference is emphatically human — 37% want a phone call, 25% want someone in person, and only 7% would choose an AI bot to sort out a problem.

Read those two sets of numbers together and a very clear guest emerges. They are happy — eager, even — to let AI do the research: the sifting, the comparing, the itinerary drafting, the "what's a good neighborhood for a family in October." They are deeply reluctant to let it do the deciding, especially where money and trust are involved. 51% told Amadeus they want a combination of AI and human expertise; only 17% trust an AI assistant on its own. Even at the moment of purchase, 89% cross-check what the AI told them before they commit. The guest isn't asking to be automated. They're asking for a fast, smart assistant on the way in — and a real person standing by for the moments that carry weight.

Why this is good news for an independent

If you run a small or independent property, this is arguably the most encouraging data to land all year — because it maps almost perfectly onto what you already are. The big chains can out-spend you on technology; they cannot out-human you. And the guest is telling us, in survey after survey, that the human is exactly the part they refuse to give up. The competitive frontier isn't "who has the most advanced AI." It's "who uses AI to clear away the friction so the human moments land better." That's a game the independent can win.

It reframes the job, too. Your guest has already done the AI-assisted homework; they don't need you to be a worse version of ChatGPT. What they need from your property is the two things the machine can't give them: a frictionless, trustworthy path through the transactional bits — a straight answer at 11pm, a booking that just works, a question resolved without a queue — and a genuinely human welcome once they arrive. When 50% of travelers expect AI assistance to cost nothing extra and only 4% think it justifies a premium, the lesson is plain: AI isn't the thing you charge for. It's table stakes now, the baseline of competence. The hospitality is still the thing you're actually selling.

What to do with a guest who's ahead of you

The instinct, when you hear "guests are two years ahead," is to panic-buy technology to catch up. Resist it. Catching up doesn't mean out-teching your guest; it means meeting them where they already are with a posture of assist, then hand off. Put an assistant on your front line that answers instantly, in the guest's language, around the clock — for availability, directions, the parking question, the standard booking — so a guest who's used to fast, capable AI doesn't hit a wall of silence the moment they reach you. And then draw the line exactly where your guests draw it: the instant a request carries judgment, money, or feeling, a human takes over, fast, with the context already in hand. That's not a compromise between AI and hospitality. It's the specification your guests just handed you.

Think of it, again, as a wingman. The machine flies the routine sorties — the repetitive questions, the after-hours messages, the first-pass planning help — and it never tires at the end of a double shift. The human is kept free and fresh for the missions that actually decide whether the guest comes back: the honeymoon that needs a real conversation, the complaint that needs a real apology, the local tip that no model will ever deliver with a wink. AI is my wingman; the human gives the welcome. Your guests have already told you they want it exactly that way.

So this week, try one small experiment. Ask the next few guests who check in whether they used AI to plan their stay — you'll be surprised how many say yes, and how specific they get. Then ask yourself where, in their journey with you, they hit a slow, human-less wall they'd have breezed past with a machine, and where, conversely, a machine would have cheapened a moment that needed a person. That map — assist here, human there — is your whole AI strategy. Your guests are two years ahead. The good news is they've also drawn you the directions.