More than nine in ten hoteliers now use AI in at least one area of their operations. Not "some hoteliers". Not "early adopters". 98%.
The conversation is no longer about whether to adopt AI – it's about where you are on the journey, and what's keeping you from going further. Those are harder questions, and they deserve better answers than most of the content out there.
In March 2026, we surveyed 500+ hoteliers to find out how AI is actually being used across the industry today. The picture that emerged is one of an industry mid-transition: broad adoption, growing confidence, and a clear divide between properties still using AI to patch operational gaps and those using it to actually drive revenue.
Here's what the data shows – and what to do about it.
Most hotels are using AI – not all of them are using it well
On average, hoteliers in our survey use AI in 11 of the 19 operational tasks we tracked. For those tasks, AI is handling around 56% of the workload.
The more a hotelier uses AI, the more optimistic they become about it – and 92% are positive about AI's role in hospitality. Hands-on experience builds trust faster than any training programme.
The properties leading on AI tend to be upscale, luxury, urban and airport hotels – segments with higher operational complexity and less room for error. Revenue management and executive leadership teams are driving most of the adoption. Front-desk teams are catching up as guest-facing use cases mature.
The barriers are real – but they shrink with experience
The two most common concerns are accuracy and data privacy. Both are legitimate. But both tend to diminish as teams build experience with specific tools in specific contexts.
The structural barriers matter more than the mindset ones. Fragmented tech stacks – systems that don't talk to each other – are the single biggest limiter of what AI can do for your property. AI can only work with what it can reach. Clean data, open APIs and connected systems aren't nice-to-haves. They're the foundation.
Three in four hoteliers have already trained their teams on AI. But training alone doesn't build confidence. What works is finding an internal AI champion – someone who experiments, learns by doing and brings the rest of the team along. One in three hotels already has this person in place. If yours doesn't, finding them is probably the most valuable thing you can do this year.
Where to start – and where to keep humans in the room
Not every task is equally ready for AI. Our survey identified four categories worth knowing.
- Safe bets for more automation right now
Rate management, data analysis, content production, review management and translation all have high adoption and high comfort levels. If you haven't already, these are the places to push further. - Guest-facing moments: hybrid works best
Check-in, concierge and guest communications are widely used, but hoteliers remain cautious about full automation here. AI handles the speed and consistency; your team handles the moments that define the experience. - Back-office: a significant untapped opportunity
Housekeeping scheduling, labour planning and procurement have lower current usage but high comfort with automation. There's a lot of time to recover here. - High-touch areas: tread carefully
Upselling, loyalty management and staff training warrant more care. The human element matters in ways that are genuinely hard to replicate.
One data point that says everything about where the line sits: 83% of experts in our panel believe hybrid models will be the industry norm by 2035. And 0% of luxury experts would fully automate the concierge role.
Four steps to get AI-ready in 2026
Step 1: Get your data house in order
Map every system you run. Identify where data is siloed or manually transferred. Clean your guest profiles. Check that your PMS has open, well-documented APIs. Without this foundation, any AI investment will underdeliver.
Step 2: Get found by AI
Your future guests are increasingly starting their hotel search in ChatGPT, Perplexity and Google's AI Overviews – not traditional search engines. 29% of Americans already use AI for travel research, and expert consensus puts the likelihood of AI chatbots fully integrating travel booking within a decade at 9.4 out of 10. Build a structured property factsheet, write answers to your top 50 guest questions, audit consistency across all your channels and make every service bookable through your own engine.
Step 3: Run your first pilot
Start small: pre-arrival email responses, review management, maintenance ticketing. Set clear guardrails. Track response time, accuracy, guest satisfaction and staff time saved over four to six weeks. Expand what works.
Step 4: Build governance and grow your team
Form a small AI working group. Define who reviews outputs and owns the roadmap. Write a simple AI policy. Find your AI champion – not a technical expert, but someone with curiosity, pragmatism and the ability to bridge technology and operations.
The ceiling is higher than most properties have reached
Hotels using Mews RMS see an average 13.7% revenue uplift per square meter compared to properties that weren't previously using a revenue management system. Revenue teams save 20-30 hours a month on average, redirected from rate adjustments to strategy and guests.
The hotels getting the most from AI aren't treating it as a software purchase. They're treating it as an organizational capability. That's the difference between AI that delivers and AI that disappoints.
The full picture – including the complete task-by-task opportunity map and a four-phase adoption roadmap – is in AI for Hotels: A Practical Guide.