AI changes discovery in travel — booking is the challenge

Travelport’s Chief Technology Officer Andrew Jordan on why AI-powered discovery doesn’t mean booking is a given

With community-led search on social media channels creating new appetites and countless options now presented instantly, AI is reshaping travel. But behind that dazzling shop window you need a platform that can turn dreams into reality—handling the real-life complexities of pricing, availability, fulfilment and compliance.

Moving people all over the world for business and leisure is difficult, and Travelport was founded on the vision to make travel less complex, provide more choice, and reduce friction. AI gives us fantastic new opportunities to deliver on that vision. In a world where all options seem to be on the table, when travelers can ask for the moon and stars and expect to get it, we’re in a strong position to help.

Social media and algorithms are shaping destination desire: 68% of users have said TikTok helps them decide what to choose in travel¹ while 59% of Gen Z travelers said they booked a trip directly due to TikTok². Those numbers will only go higher. Travelers increasingly expect AI to understand intent, to narrow choices, explain trade-offs, and accelerate the move from inspiration to booking. Statista reported in 2025 that 40% of consumers worldwide were already using an AI-based tool like ChatGPT for travel planning.³ But there remains a significant gap between inspiration and transaction.

I should perhaps differentiate between “untouched transactions” that can be fully automated, customer queries and simple bookings, for which AI is great. And “touched transactions” that need human intervention. There’ll still be a desire for more traditional high-touch travel providers—bespoke or luxury travel, and multi-stop trips with complex itineraries, handled by specialist agencies. And there will still be a need for a human in the loop if things go wrong (delays, cancellations).

Discovery has changed, booking hasn’t

At least not yet. That gap is the problem we are solving.

Modern travel retail has traditionally been destination- and availability-based: where you want to go, when, how many people. With AI enabling what appears to be a limitless range of possibilities, there is now real potential for true freedom and personalization. AI excels in the pre-booking stage (discovery, planning, itinerary creation). But travelers, or their agents, also need to move swiftly and confidently to the next step, booking and payment.

AI-driven discovery today remains imprecise and less governed compared to traditional availability-based booking systems. AI is probabilistic; outputs depend on training data and can be inaccurate or prone to “hallucinations”. This contrasts with the core requirement in travel booking: precision and certainty, especially for payments and logistics—and with risks including incorrect transactions, invalid itineraries, and duty of care obligations in corporate travel.

To date, the travel industry has struggled to properly connect discovery to booking, but examples do exist. Travelport customer Sembo uses a smart algorithmic approach to create holiday packages based on natural language input: “I want two weeks in Florida for a family of four, a road-trip with six destinations, flying into Orlando, with X budget.” Once an itinerary is presented, a button invites the customer to buy that trip immediately. Because Sembo is pulling live availability from Travelport, the trip is fully bookable in real time. This goes beyond research—it directly feeds into a transaction.

While AI tools can already generate personalized recommendations from natural language input, the travel industry and its systems have not been structured to handle these complex, open-ended queries at scale. Legacy technology standards like XML and JSON limit efficiency compared to modern machine-readable systems. AI-driven travel retail requires more efficient query handling, improved data processing, fewer unnecessary permutations, and more relevant results supported by intelligent content curation.

And change is coming. Travelport recently announced a partnership with AI builder and technology services provider Cognizant and Anthropic, the company behind Claude, to connect systems that can reason and plan with platforms that can transact in the real world. A new AI travel ecosystem is coming.

For now, scale, cost and infrastructure can be significant challenges; AI requires substantial computational resources delivered cost-effectively. Trust is another barrier, though less so among younger travelers. A global survey by G2⁴ found that 51% of people aged 18–34 trust AI; 48% of those aged 35–54 trust AI; and only 38% aged 55+ trust AI. Trust will improve as AI becomes as familiar and accepted as today’s messaging tools.

Infrastructure must match the shift

Travel platforms must provide transactional precision, access to reliable inventory and availability, compliance, and the ability to manage disruption and cancellations—capabilities that AI alone cannot deliver. Travelport’s strength lies in providing this reliability and certainty at scale, across both leisure and corporate travel, underpinned by unparalleled travel data and industry experience.

A strong AI-ready platform enables you to access and use complex real-time inventory, combining multiple options from countless data sources, accepting payments, supporting ticketing, settlement, and compliance. On their side, travel sellers need to increasingly focus on adapting their culture, systems and processes to AI-influenced behaviors, to support a seamless transition from inspiration to booking. And this depends on having the right infrastructure.

Today, AI is most effective in supporting discovery, enhancing personalization, and improving efficiency, but existing systems and human oversight remain essential for final transactions. The connection between inspiration and booking is evolving fast, with even smarter ultra-personalization and truly hybrid AI/human experiences on the way. 

Travelport is setting the pace in an AI-first distribution landscape where speed, content normalization, and real-time connectivity are paramount. We’re doing this by delivering a scalable intelligent platform, expanding our content powerhouse, and enabling a truly AI-driven experience. Those three pillars may appear simple but, like a swan gliding across water, there’s an awful lot of paddling below the surface—powered by a $50 million investment in the next stage of our company’s growth. 

This is the first in a blog series exploring how AI is transforming travel—and how the businesses that succeed will be those with the infrastructure, platforms and expertise to make AI work at scale. The next blog looks at how AI can stall when the stakes go up.