Skift is taking its Data + AI Summit to London for the first time on Oct 6, 2026, betting that Europe — with its multiple languages, brands, booking systems and regulators — is where travel AI either proves it can convert convincing answers into completed, changed and paid-for bookings, or stalls.

The gap between recommendation and transaction frames the event's agenda. An AI assistant can suggest a Barcelona hotel to a London traveler in seconds, but closing that booking requires handling availability, room types, payment and cancellation terms. Modifying it later adds another layer of complexity. Add a connecting flight, a corporate travel policy or a mid-conversation language switch, and the distance between a polished demo and a deployable service widens.

For sellers of travel, the stakes are distributional. If more travelers delegate comparison shopping to an assistant, suppliers and intermediaries need to understand how they enter that consideration set — and what happens to the direct traffic they previously received.

Metasearch Chiefs Confront Delegated Search

Skyscanner CEO Bryan Batista and KAYAK CEO Peer Bueller, whose businesses were built on helping people compare options, face the core question head-on: what happens to comparison shopping when the traveler stops doing the comparing? For metasearch, the answer determines whether referral economics survive the shift to agentic booking.

Hotels face a related but distinct problem. A recommendation only matters if it produces business on acceptable terms. Joe Pettigrew of L+R and Carlo Del Mistro of Stiplo will press the visibility question: what can a property actually influence, which data needs fixing, and how does a hotel distinguish a genuinely useful new booking channel from another marketing expense whose results resist measurement?

Borders sharpen the problem. A hotel must be legible to customers searching from different countries, in different languages, and with different assumptions about what matters in a stay.

One Group, Many Operating Environments

Inside travel companies, the challenge compounds. An airline or hotel group may share a single AI ambition, but its brands, systems and teams rarely start from the same place. IAG's chief AI scientist, Ben Dias, and Accor's data science and AI leader, Nicolas Maynard, join the program to address where a common approach helps, what must remain brand- or market-specific, and how much of the work is AI at all versus making information usable across the business.

Execution gets its own session. Strawberry's Kari Anna Fiskvik and Amadeus's Benjamin Cany bring the perspective of what it takes to get an AI product live. OAG CEO Filip Filipov carries the data question: a decision-support service needs information accurate enough, current enough and usable at the exact moment a traveler decides.

US companies face the same problems when they operate internationally, but Europe makes them impossible to postpone — cross-market operation is the proposition, not an edge case.

Regulation Becomes Part of the Product

Europe adds an explicit AI regulatory layer. The EU's AI Act transparency requirements began applying in August, with major high-risk obligations following later. The UK is charting a different course through its existing regulators. For any business serving customers across both regimes, governance must sit inside deployment planning rather than around it.

The compliance questions extend past whether a chatbot identifies itself. What customer data can a service use? What authority has the traveler granted it? When should a machine hand a decision to a person? Sheena Varma of American Express Global Business Travel brings privacy, risk and compliance experience to a discussion Skift argues belongs alongside product and commercial decisions, not apart from them.

Incumbents Hold an Asset — If They Can Use It

The summit's underlying thesis favors incumbents that move. Established travel companies know how bookings work, where the exceptions occur and what customers need when plans collapse. That operational knowledge could appreciate in value as AI absorbs more of the journey — but only for companies that make it machine-usable.

Investors and entrepreneurs examining what deserves funding join the London conversation, and the program is aimed squarely at teams building AI into their 2027 plans. The argument, in short: the companies that solve AI's hardest cross-border problems in Europe will shape how travel gets sold everywhere else.