B y the end of 2026, the travel industry may have survived hundreds of articles and con- ference panels addressing The Future of AI in Travel . There will be more. There are always more. Still, beneath the noise, something useful is happening. Artificial intelligence is moving from novelty to infrastructure. Travellers are using it to dream, search, compare, translate and plan. Hotels are using it to manage revenue, staffing, energy and guest messaging. Airlines are testing it across disruption management, customer service and operations. Destinations are using data to manage flows, capacity and demand. The industry survived the internet, OTAs, mobile, metasearch, guest reviews, contactless check-in and the era of people filming breakfast buffets for TikTok. It can survive agentic AI. The better debate is what to do with it. The “AI versus humans” argument is tired. The useful version is simpler: which decisions can be automated, which moments need people, and which content is structured well enough for a machine to trust? THE BOT GETS A TO-DO LIST Agentic AI is the phrase everyone is circling. In plain English, it means AI that can do more than answer a question. It can follow an objective, use tools, call APIs, remember context, make decisions within rules and complete a task. A chatbot suggests three hotels. An agentic system checks availability, compares rates, applies loyalty preferences, books the room, changes the airport transfer when the flight is delayed and sends the invoice to expenses. Suddenly, the bot has a clipboard. That is why the fuss matters. Phocuswright says more than 60% of travel businesses it surveyed are already experimenting with or scaling agentic AI, although only 6% are scaling it across multiple domains and 22% have started scaling selected areas. Tech budgets are also rising, with 62% of travel companies expecting increases. The reality check is just as important. McKinsey’s travel executive research found 90% are using generative AI in some form, but only 2% say agentic AI is in widespread use. Travellers are cautious too: more than 90% have some confidence in AI travel information, but only 2% are willing to hand over full autonomy for AI to make or modify bookings without human oversight. • TECHNOLOGY & INNOVATION
After the AI panic
Everyone is talking about agentic AI. Fair enough. But travel’s 2026 tech story is less science fiction than operating discipline: better content, cleaner data, smarter workflows and a traveller who expects the journey to work like the rest of their digital life.
Above: No need to panic: travel can survive agentic AI
So, yes, agentic AI is coming. No, it is not about to run your hotel while you go for lunch. HOTELS: FEED THE MACHINE Hotels are right to pay attention though. Agentic AI changes how travellers may discover, compare and choose accommodation. Search has already started to move from “show me 20 blue links” to “give me three good options and tell me why”. Carrying the slogan, “Your trip. Planned in minutes”, Layla.ai is a good example of what that shift looks like in the wild. The Berlin-based AI travel planner describes itself as an “AI travel agent and trip planner”, creating personalised itineraries covering flights, hotels, activities
and dining in minutes. Users share dates, destination ideas, budget and travel style, and Layla builds a day-by-day plan using live pricing and availability. Back in March, the company said it had already processed more than 30 million travel messages from over five million users, generating more than two million trips and more than US$1 billion in planned trip value. More tellingly, 40% of users now start without a destination, asking by mood, need or intent rather than typing in a place name. For hotels and destinations, that is the point: if AI is helping choose the destination before the traveller has chosen one, content must be built for inspiration as well as conversion.That puts hotel content back at the centre of the commercial strategy.
68 | ATM YEARBOOK 2026
Made with FlippingBook Digital Proposal Creator