Hotels, tour operators and travel agencies are rushing to launch proprietary online tools and loyalty schemes to fend off future competition from autonomous bots, capable of arranging travel for their users.

A survey last year by OnePoll for travel aggregator Skyscanner found that, when asked about future travel plans, 38 per cent of some 22,000 respondents said they would use AI tools to research a destination, while 33 per cent said they would use them to plan a holiday itinerary.

“This is not something futuristic,” says Jean-François Guilmard, chief data and AI officer at French hotel group Accor. Travellers’ holiday planning and booking processes are “already changing drastically”, he adds.

So far, holidaymakers are mainly using AI to search for destinations and compare accommodation, while making reservations through online travel agents such as Expedia or Booking.com or hotels’ systems. But Accor, which owns brands including Sofitel, Mercure, Fairmont and Raffles, is “already preparing for the next step [which is] the agentic step,” says Guilmard.

That push includes investing in integrated apps that can facilitate bookings and payments directly within large language models, such as ChatGPT, as well as promoting a loyalty programme.

More from AI in Practice

Budget-busting AI bills; transforming pharma; AI and trust in air traffic control; factory robots on the march; AI and mental health; astronomical implications; rewriting gaming rules

Bank of America analyst Justin Post noted in a report in January that loyalty ecosystems and personalised rewards would be crucial for travel operators seeking to stay at the forefront of customers’ minds amid the rise of AI. Guests who want to let an AI agent book for them will want confidence they will be satisfied, says Guilmard. “If you trust a loyalty programme, that will make a difference.”

Optimists in the hotel sector hope AI agents will provide a cheaper alternative to online travel agents such as Booking.com and Expedia, whose fees have been a source of frustration to some. But others fear that AI tools could divert business away from big industry names by reducing brand recognition.

The intermediaries might be under the greatest pressure if consumers turn to “agentic” AI chatbots capable of arranging travel without using their services. Shares in Booking Holdings, which owns platforms including Booking.com and Priceline, have dropped 15 per cent in the year to date, while rival Expedia has fallen 7 per cent.

Surviving that threat means “meet[ing] customers where they are,” says Skyscanner chief executive Bryan Batista. In addition to building an app within the ChatGPT marketplace, Skyscanner has recently enabled “conversational search” on its website to cater to AI-native customers.

A user might, for example, search for a “cheap but sunny hiking holiday from London for a week in November” and receive recommendations that include insights on the average price and daily temperature for each suggested destination, such as Tenerife and Madeira. Batista says this could make travellers more open to new destinations, thus spreading out demand and reducing congestion in the busiest hotspots.

Batista also argues that brand recognition would matter even more in an AI-driven world. Fears of AI hallucinations mean prospective holidaymakers are “asking more and more for trustworthy guidance grounded in real experience and data”, he says.

Meanwhile, he says that AI-powered coding tools are helping companies react to customer demand and build features around specific events. Skyscanner took less than a week, using agentic coding tools, to build its Football Flight Finder, which directs people to flights to follow teams in the World Cup, he says.

Personalisation is also increasingly critical to companies’ relationships with their customers: “Travellers now expect us to remember them [and] we have no excuse to not do that because the LLMs do it effectively,” says Batista.

For Guilmard, the rise of AI is not a threat, but simply a different channel through which Accor can sell hotel rooms. But he acknowledges that customers’ costs will change and says there is “no blueprint” on how that will work out. “What was quite well known with classical traditional search — search engine optimisation and so on — all of that is far more blurry now with AI.”

Other entrepreneurs, meanwhile, are hoping to make money by building AI tools to help shape customers’ booking processes. Martin Perez, for example, believes that new technologies such as his Singapore-based start-up BookYolo will help consumers surface the best travel deals and advice, reducing what he calls “post-booking regret”.

BookYolo started from personal frustration. Perez had booked accommodation to work remotely overseas. His own scan of the property’s reviews raised no alarm bells but, when he arrived, there were two loud construction projects nearby, making it “almost impossible” to focus on his work.

The entrepreneur says he would have chosen an alternative if he had been able to efficiently filter the thousands of reviews, other data and news reports online. “It’s too time-consuming for a person. It’s too much information,” he says. “But you can use the power of AI to really dig into what’s relevant and make the best decision about the listing based on your personal needs.”