AI is coming for everything in hotels – except what you choose to keep human

For years, getting your hotel found online meant one thing: winning at Google. Now a growing number of guests skip the search bar entirely and ask ChatGPT where to stay. For hotel owners, it’s a fresh opportunity to reach new guests. 

Gilad Berenstein has been thinking about this longer than most. He founded Utrip, an early AI travel company that helped more than 200 million travelers plan trips before it was sold. Today he runs Brook Bay Capital, investing in AI across travel and hospitality. 

On the latest episode of Matt Talks Hospitality, Gilad sat down with Mews CEO, Matt Welle, to talk through the AI use cases hoteliers can act on now, and the ones that wait further down the road. 

Start at your own desk 

Plenty of hoteliers still feel lost when it comes to how to actually use AI to help their hotel. Gilad’s answer had nothing to do with strategy decks or vendor shortlists. 

“When you show up at your computer in the morning, you might open an email, a CRM, Salesforce, whatever. You should also be opening your AI, and you should have it there with you all day,” he said. 

Daily use is where small opportunities surface, some of which can grow into big opportunities. “Until you build personal competence, there's no real reason to talk about business strategy,” he added. Once that habit is in place, the next job is creating the kind of content a property needs to be discovered in an AI-led search landscape. 

Your inbox is already training data 

The shift Gilad watches most closely is the rise of AI swarms: groups of agents that are each fairly simple on their own but get far smarter when they talk to each other. Hotels are starting at ground level, connecting a few agents that handle commissions, customer service and operations. 

A good example is the concierge. Take every guest email asking what to do in Rome or Venice, plus every reply your concierge has written by hand – that's your perfect training data. Add an agent for predicting no-shows and another for chasing commission, and once you connect them, the knowledge starts to compound. 

Most hoteliers already sit on this data gold mine; AI can actually make use of it. The goal isn't fewer people. As Gilad said, “I'm not suggesting that you replace your human concierge. I'm suggesting that you superpower the human concierge.” 

The fully connected agentic version – where a guest's own agent plans dinner with the hotel concierge agent – is “an end state that does not really exist yet.” His estimate for getting there: at most a couple of years, and probably months. 

Build what makes you different 

Now that almost anyone can spin up a chatbot, the question isn't whether you can build something. It's whether you should. 

Start by seeking out the repetitive tasks. “Whatever your team is doing over and over and over again right now clearly is not accounted for by your system,” Gilad said. That's the low-hanging fruit, and the right place to begin. 

Beyond that, it comes down to what sets your property apart. Accounting works the same way in almost every hotel and the same goes for Wi-Fi: “No customer's ever going to choose your hotel because you have a better Wi-Fi login experience than someone else.”  

That’s not where you should invest your time and resource – or where you should focus your AI building. But if you run a wellness hotel or a sustainability-led property, that's where building your own tech starts to make sense, because it sharpens the very thing guests choose you for. 

Where AI is earning its keep 

As of right now, AI has a limited impact on the guest-facing side of hospitality. Few guests check out of a hotel thinking AI made their stay dramatically better. But where it does already make a big difference is behind the scenes. 

Gilad talked about Shashi Hotel in San Francisco, a new build with sensors in the walls tracking noise, heat, dampness and water. A leak can now be caught in one bathroom wall before it spreads through the tower, which has paid off quickly on property insurance. 

At Atlantis The Royal in Dubai, staff already carry location beacons for security. The hotel uses that same data to tell guests roughly how long a request will take to arrive. It's a small thing, but a genuine piece of luxury. The time it takes to deliver it isn’t the point – it’s the knowing. 

The €30 trip upgrade 

Gilad’s daughter loves princesses. In his picture of a truly personalized family trip to Paris, the itinerary becomes a princess-themed video on the drive to the airport. Toys are waiting by the hotel bathtub when the family arrives. The next day's walking tour guide comes armed with fairy tales to keep the kids hooked, and a box of gluten-free cookies turns up so his wife can enjoy a Parisian dessert. 

The added hard cost of all that, by his estimate: about €30. What he'd pay on top? “I think it's almost certainly €300. In a luxury setting, it might be €3,000.” 

The same logic works at a much smaller scale. Gilad doesn't drink much, but he really loves chocolate. Swap the welcome bottle of wine, which costs $30 or $40 even at wholesale, for a beautiful $8 bar of local chocolate, and the math changes in the hotel's favor. “Not only did you spend less than 20% as much, I feel seen and cared for and engaged with.” 

What stays human 

For all the talk of agents, sensors and robots, Gilad’s vision of a hotel is surprisingly people centric. “I don't want fewer humans working in your hotel. I want fewer humans locked in the back hallway, in the back office, in the laundry room,” he said. 

He described a recent ten-minute check-in at a high-end hotel in Asia that was entirely transactional. Late at night, the whole thing could have been automated. Any other time, it could have been hospitality: the hotel had a full spa floor that a team member could have shown him straight away. 

Guest intelligence tools like Abra point the same way. A guest tells the bartender they love mangoes, the bartender logs it, and the next morning housekeeping is prompted to swap the strawberries for a mango. The tech handles the memory. The person still makes the moment. 

“AI is coming for everything except for those things that we choose to keep in the human domain,” Gilad said. For hospitality, that means training staff to make guests feel welcome, give them a sense of belonging and help them be true to both themselves and the place they're visiting. 

In other words: AI can take care of the admin side of things, so that humans can handle the hospitality.  

Listen to the full Matt Talks Hospitality conversation between Matt and Gilad – and many more like it – wherever you listen to your podcasts. 

Watch the episode