Hotels keep telling researchers they don’t trust AI to set their room rates. I think that’s the wrong worry. The bigger problem is that a huge share of hotels don’t have a pricing system worth trusting in the first place, human or machine. And the guests are about to send AI shoppers to compare those prices for them.
The trust debate makes for a good headline. The real story is what sits underneath it.
What the new research says
RoomPriceGenie’s 2026 State of Revenue Management Report, reported by Hospitality Net, surveyed 450 senior revenue decision-makers across the UK, US, Canada, Mexico, DACH, Spain, Australia and New Zealand.
The headline numbers:
- 53% of hotels already use AI to help set room prices, and 16% use it extensively.
- 77% say they trust AI to at least some degree.
- Only 10% would let it make most pricing decisions on its own. Most still check its recommendations by hand.
That is the bit everyone will quote. Now read the next part.
- 43% of hotels have no revenue management system at all.
- 14% price on instinct or fixed seasonal rates.
- 29% use spreadsheets with limited automation.
- Only 30% say gathering the pricing information they need is quick or easy.
- Only 24% give a dedicated revenue manager primary responsibility for pricing.
RoomPriceGenie sells pricing software, so of course it wants hotels to buy more of it. Keep that in mind. But the numbers still tell a clear story.
Keeping a human in the loop is not the problem
I’ll say something that may sound odd from someone who talks about AI all day: the 10% figure doesn’t bother me. A revenue manager who checks the machine before pressing the button is doing their job. Pricing a hotel is not just maths. It is strategy, brand position, group business, the conference next door that the data hasn’t spotted yet.
RoomPriceGenie’s CEO, Chas Scarantino, made the right point. Hoteliers, he said, “need to understand why a price is being recommended, make sure it fits their strategy, and feel confident acting on it.” That is an explanation problem for the vendors to fix, not a character flaw in hoteliers.
So let’s stop treating “hotels don’t trust AI” as the scandal. It isn’t.
The scandal is the 43%
Almost half the hotels in this survey are pricing by gut, by last year’s seasonal grid, or by a spreadsheet someone built in 2019. And more than two in three say getting the information they need to price is not quick or easy.
That was survivable when a guest looked at a handful of options on a couple of websites. Scarantino again: “guests compare dozens of options before they book, so a wrong price costs real money, fast.”
Now think about who does the comparing next. Not a tired guest with ten tabs open. An AI assistant that can check far more options than any person would bother to.
AI shoppers will spot a lazy price
This is where the pricing story meets the AI discovery story. At a Skift Global Forum roundtable presented by Merkle, Adobe data shared by the panel showed AI referral traffic up a reported 3,500% between May 2025 and May 2026. More importantly, AI-referred visitors now convert at a rate only about 1% behind non-AI traffic.
Read that carefully. The people arriving from AI are no longer just browsing. They are arriving ready to book.
I’ve argued before that the small share of traffic AI sends hotels is the wrong number to watch. Here is another reason. If an AI assistant shortlists five hotels and yours is priced off a fixed seasonal rate while the city is half empty, you don’t look “premium”. You look wrong. And the assistant has no loyalty to you. It will simply recommend the hotel next door.
A human revenue manager can get away with gut feel for a while. A machine comparing rates in seconds will not be so forgiving.
Fix the basics before you argue about autonomy
This is the same pattern I keep seeing across hospitality. Hotels adopt AI, but the spreadsheets don’t leave. Teams buy ChatGPT but not an AI strategy. And now, it seems, hotels debate how far to trust AI pricing while four in ten have no real pricing system behind them.
If I ran a hotel, my order of work would be simple:
- Get a real pricing system. Any system that reacts to demand beats a seasonal grid.
- Keep the human, change the job. Let people set the strategy and the guardrails. Let the software handle the hundreds of small daily moves.
- Make your rates and offers easy for machines to read. If AI assistants can’t see your price clearly and accurately, they will use someone else’s.
- Ask your vendor to explain itself. If a tool can’t tell you why it picked a price, the trust gap is the vendor’s fault, not yours.
The real question
The industry is asking, “Should we let AI set our prices?” I think the better question is, “Would our prices survive being compared by AI?”
For a lot of hotels, the honest answer today is no. And that has nothing to do with trust.
Where do you stand? Is the 10% figure a sign of healthy caution, or a sign hotels are protecting habits they should drop? I’d love to hear from revenue managers and hotel marketers on this one.
Want more takes like this? I share what AI is really doing to travel search, booking and marketing every week. Subscribe to my LinkedIn newsletter.