For many travelers, the process of planning their trips is a grueling slog through endless price comparisons and messy browser tabs. Increasingly, they are outsourcing the task to artificial intelligence.
About 91% of global travelers rely on AI travel planners, according to an annual survey of 11,000 global users by travel platform Klook.
Travelers use the technology for various reasons. Some turn to AI to figure out what they even want from the trip, and others to find the best possible deals that suit their needs, Klook’s research revealed.
The appeal partly reflects a “do it yourself culture, feeling the ownership and pride of like building something yourself,” said Leigh Rowan of travel agency Savanti Travel.
But trust remains a challenge. A separate report by Booking.com on consumer attitudes towards AI in everyday life and travel found that 91% of respondents continue to have concerns about AI, with only 35% fully trusting its outputs.
Over the shoulder view of young Asian woman using smartphone, checking digital flight ticket and boarding pass on device screen while packing a suitcase on bed for a trip. Planning for travel. Travel and vacation concept
D3sign | Moment | Getty Images
‘Hallucinations’ hurdle
One major obstacle is accuracy. AI tools, which are built around large language models, are known to produce ‘hallucinations’, a phenomenon where false information is presented as fact.
Shyn Yee Ho, a director at tourism consultancy Horwath HTL, said her experience with AI was largely devoid of hallucinations.
As a self-proclaimed “heavy user” of LLMs, Ho relied on suggestions from ChatGPT to find hotels within her budget and destinations that matched her interests while planning for her six-month solo sabbatical trip. Those recommendations, she said, were “very clear and good.”
While Ho’s experience with AI has been positive, the reliability of AI continues to be a concern for many travelers.
Savanti Travel’s Rowan recalled a client, who was in Paris on a business trip, arriving late for an appointment after ChatGPT suggested a route that failed to account for road closures due to construction. What should have been a 10-minute transfer turned into a 45-minute journey.
“They seem like they’re edge cases, but they’re actually very common,” Rowan said.
The AI paradox
Industry experts say AI may also reshape which destinations travelers see. Smaller businesses may be particularly affected because they lack the expertise or digital presence needed to appear in search results.
Older properties, especially independent ones, or even those in developing countries, “will struggle even more because they don’t have the expertise, they don’t have the guidance,” Horwath HTL’s Ho said.
“It is a shame,” she said, “because arguably… they need the demand more than ever, right?”
On the flip side, well-publicized tourist destinations might suffer from the consequences of over-tourism in AI’s current iterations.
AI tools should be used to introduce offbeat locations, to spread out demand so there isn’t that risk of over tourism, Guy Llewellyn, assistant professor at EHL Hospitality Business School Singapore, said. Many AI systems are being trained on “the top 10 lists” of places to recommend, which is a bit of a paradox, he added.
Crowds of tourists on the street near Kyoto’s Kiyomizu-dera temple in April 2019.
Nicolas Datiche | Lightrocket | Getty Images
AI tools may also struggle with real-world nuances that experienced travel planners account for. These include factoring seasonal weather when suggesting outdoor experiences and travel fatigue after long flights, Rowan said.
“Multi-generational movement issues, you know, outright allergies, disabilities, intolerances, those things AI really can’t manage very well,” he said.
It’s also not able to help you when things go wrong, he added. When the war in the Middle East started, numerous travelers were stranded after the airspace in the region was shut down. AI isn’t going to get you to the top of the queue for the next repatriation flight, he said.
The road ahead
However, as AI models improve and businesses become more savvy at integrating new tools, Llewellyn is hopeful about their utility.
“As we’re getting more integrated with the AI, they might actually end up having more up-to-date data than the traditional travel agent,” he added.
The hospitality industry should prioritize structuring and opening its data so AI can access it, Llewellyn said.
“It doesn’t have to be customer-facing; it can just purely be on the back end through an API call,” he added, referring to “application programming interface,” or a gateway that allows outsiders to use a company’s data for different purposes.
As long as the AI had access to it, he said, it would be able to provide factual information to the end user and reduce overall hallucinations.
Despite the information gaps and deficiencies of existing AI models, the hospitality industry is pressing ahead with the rollout of more smart tools. Booking.com, for example, has implemented various AI-powered solutions with OpenAI.
Llewellyn predicts that the digitization of information will continue to proliferate as more AI companies enter the market.
Travelers should benefit from this, as it should reduce the frequency of hallucinations and improve the overall user experience.
“AI planners are going to happen,” Llewellyn said. “The first few iterations are going to be slow. They’re going to have some issues, but it’s going to be a really impactful thing for the industry.”