World Tourism Day 2026: How AI Is Quietly Rewriting The Way India Travels

Every September 27, the United Nations pauses to ask what tourism means to the world. This year, the question arrives with an answer already built into the occasion. World Tourism Day 2026, hosted for the first time by El Salvador, carries the theme “Digital Agenda and Artificial Intelligence to Redesign Tourism,” with sessions built around harnessing technology to reshape how visitors are served, and destinations are discovered and managed.

The industry’s own numbers back up why this theme was chosen now rather than five years ago. In a sixth annual Travel Outlook Survey, a third of respondents said they were likely to use AI tools such as ChatGPT, Gemini, Copilot, or Claude to help plan travel this year. Among that group, three in four expected to lean on AI for recommendations and seven in ten for itinerary planning, but only about one in ten expected to trust it with the actual booking. A separate 2026 hospitality study found that once travellers try AI for trip planning, the habit sticks fast: 63 per cent of those who had used it said they now relied on it for most or every trip, and 96 per cent said they would use it again. Put together, the data describes an industry where AI has already taken over the browsing tab but hasn’t yet been handed the “confirm booking” button.

To understand that gap, and where it is closing fastest, Outlook Traveller spoke with three people building AI into travel from very different vantage points: Vatsal Desai, founder and CEO of the luxury travel platform Vaultfy; Vinod Kumar Sah, co-founder and CTO of COTRAV, a concierge-led corporate travel and hospitality management platform; and Hari Ganapathy, CEO and co-founder of the itinerary company Pickyourtrail. Their answers, taken together, sketch out an industry that has quietly agreed on where the line sits, even if none of them planned it that way.

From Search Engine To Conversation

The most basic change all three describe is what a traveller is now allowed to ask for. Planning a trip used to mean, in Desai’s words, “visiting multiple websites, comparing options, contacting agents, and coordinating different providers.” What AI has done, he argued, is collapse that into something closer to a conversation: “AI is fundamentally changing travel from a search problem into a conversation,” he said, pointing out that a request as layered as a weekend in Paris, a table at an exceptional restaurant, and a private flight home no longer has to be handled as three separate transactions. “AI can understand them as one experience.” That thinking sits behind Alfred, Vaultfy’s AI concierge, and behind a conviction Desai returns to more than once: “The real luxury isn’t having unlimited options. It’s having your time back.”

Ganapathy sees the same shift but is more precise about where it stops. “The shift hasn’t happened where most people assume it has. It’s happened in discovery, not in the actual booking,” he said. Almost every recent trip has had AI touch some part of it, comparing hotels, building a rough itinerary, working out what’s worth seeing, and “that part has moved fast.” The booking itself has lagged for a reason he thinks is entirely sound: “a flight or a hotel booking is real money and real consequences if it goes wrong.” He does expect the simpler transactions to catch up soon. “The structured, high-confidence parts of travel, flights, and branded hotel inventory are becoming genuinely agent-bookable, and I’d expect that to be mainstream within the next year, maybe sooner.” Anything with more moving parts, multi-city routing, visas, anything expensive to get wrong, will keep AI in a supporting role: “It’s doing the comparing and the coordinating, while a person still signs off before money actually moves.”

Sah’s version of this shift is shaped by a constraint the other two don’t have to plan around: company policy. At COTRAV, AI is used to progressively learn a corporate traveller’s preferences, airlines, timings, seats, meals, and hotels, so that “instead of employees repeatedly communicating the same preferences for every trip, this information can increasingly be used to present more relevant booking options from the beginning.” But he’s careful to note that preference alone can’t drive a corporate booking. Recommendations “also need to account for company travel policies, eligibility, budgets, and approval requirements,” which is why COTRAV’s approach combines personalisation with organisational controls rather than optimising for the traveller’s taste alone. The upside he sees isn’t speed for its own sake but redistributed attention: “AI can take on more repetitive information processing, search, and recommendations, allowing our teams to spend more time on exceptions, negotiations and complex traveller requirements.”

Where A Perfect-Looking Itinerary Meets An Unbookable Table

Ask any of the three where AI itinerary planning actually goes wrong, and the answer converges on the same idea: the plan on the screen and the trip on the ground are not the same object.

For Desai, the failure mode is the fact that AI has no way of confirming. “AI can recommend a Michelin-starred restaurant without knowing whether a table is available. It can suggest a private yacht without confirming its location or availability, or create an ambitious itinerary that overlooks travel times, local restrictions, and seasonal conditions,” he said. In the high-value travel Vaultfy handles, that gap is expensive to ignore, so the company treats AI as the layer that understands and coordinates a request while “verified availability and human expertise ensure the experience can actually happen.” As he put it, “AI can create the perfect itinerary on paper. The real value lies in making it happen in the real world.”

Ganapathy framed the same gap as a confidence problem baked into how these models work. “An AI will confidently send you snorkelling at a resort whose house reef is mediocre, route you into a destination in exactly the wrong season, or string together transfers that look fine on a screen and simply don’t connect in real life,” he said. “It plans with total confidence and zero experience of ever having actually gone anywhere.” He also pointed to something less obvious: most travellers struggle to articulate what they actually want. “Someone planning a honeymoon doesn’t say ‘anywhere nice’ because they haven’t thought about it; they say it because what they actually want is hard to put into words. AI is very good at answering the question you asked. It’s not yet good at understanding the question you meant.”

Sah’s concern is less about imagination and more about the freshness of data. “AI may not always have access to the latest pricing, availability, operating hours, location changes, local disruptions, or offers,” he said, which is why COTRAV treats an AI-built plan “as a very good starting point, but not automatically as the final source of truth for a trip.” Even a genuinely good itinerary can leave a traveller stranded across platforms, needing to “move between different platforms to book the flights, hotels, cabs, and other components of the journey,” so the plan “may look seamless on paper, while the actual execution remains fragmented.” His conclusion doubled as a working philosophy: “AI can be an excellent co-pilot for travel planning, but it should not necessarily be treated as the single source of truth for a journey.”

A Companion For The Journey, And A Line Around What’s Real

Where all three get genuinely animated is what happens once the trip has actually started, and what AI should never be allowed to fake.

Desai described the next phase as a move “from an itinerary planner to a companion that understands you throughout your journey,” one that already knows a traveller’s dietary preferences, favourite experiences, and family requirements well enough to adjust a plan the moment a flight is delayed, or produce a private dining reservation on request. “The technology should adapt to the traveller, rather than expecting the traveller to adapt to technology,” he said.

Ganapathy agreed that the more interesting opportunity sat after departure rather than before it. “The bigger opportunity is on the ground, flagging a cancelled connection before a traveller even reaches the gate, resending a voucher at 11 pm, answering the dozens of small, practical questions that come up mid-trip,” he said. But he drew a firm boundary at the moment something actually breaks: “If a connection breaks at midnight in an unfamiliar city, nobody wants a chatbot; they want a person they can call. The future, I think, is AI handling everything that leads up to that moment, so a human is only needed for something that actually matters.” Sah’s team is already building towards that connective layer at COTRAV, where the goal is to link a flight change to “other parts of the journey, such as hotel arrangements and airport transfers, instead of treating each booking independently.”

The same instinct shows up when the conversation turns to AI-generated travel photography, images of beaches and skylines nobody in the picture has actually stood in front of. Desai and Ganapathy land on almost identical ground, from opposite ends of the industry. “There is nothing inherently wrong with using AI to visualise a destination or inspire a future journey, provided it is clearly presented as AI-generated,” Desai said. “The problem begins when fictional experiences are represented as real ones, particularly by influencers or brands. AI should inspire people to experience the world, not convince others they’ve experienced a world they never visited.” Ganapathy called a fabricated travel photo “the visual equivalent of an itinerary you’ll never actually travel: impressive, confident, and disconnected from anything real,” adding that using it to manufacture the appearance of an experience “is a different thing, and not one I think travel should get comfortable with. Our business exists because a holiday is a real thing that happens to a real person in a real place. Better tools don’t change that.”

That is, in the end, the quiet agreement running under this year’s World Tourism Day theme. AI has already become the fastest-growing layer in how a trip is imagined, compared, and shaped. What none of these three founders is willing to hand it, at least not yet, is the moment a booking becomes real money, or a memory becomes a claim about a place someone never went.

FAQs