Planning a holiday used to mean starting somewhere: a search engine, a travel magazine, a booking site. You entered a destination, browsed options, built a mental model from what the platform showed you. The platform had leverage — it decided what to surface, what to promote, and in what order. That model ran largely unchanged for two decades. It is now shifting faster than most of the travel industry prepared for.
AI tools — general-purpose chatbots and increasingly specialised travel assistants — have become a legitimate first stop for trip research. People are asking for complete itineraries, comparing destinations by interests and budget constraints the platforms never offered as filters, and arriving at booking decisions that are largely already made. The discovery phase, which used to happen inside brand-owned channels and interfaces, is increasingly happening in conversations elsewhere.
For travel companies, this is uncomfortable. The traditional funnel assumed travellers arrived uncertain and were guided toward a decision by the platform’s interface. That interface was also the primary advertising surface — search rankings, display placement, promoted listings. When travellers show up with a specific request rather than a general query, the whole architecture of influence that platforms were built on starts to erode.
The response from established players has been mixed. Some OTAs and search products have integrated AI summaries into their interfaces. Google has pushed AI overviews into travel queries. But the more significant shift is that the primary AI tools most travellers use are not owned by travel brands at all. The planning conversation is happening in a space the industry did not build and largely cannot control.
This creates a real tension for travellers too. AI trip planning is genuinely useful — fast, personalised, capable of synthesising information across dozens of destinations — but it also has a well-documented reliability problem. Chatbots invent restaurants that don’t exist, reference hotel amenities that have changed, and produce itineraries that look reasonable but contain errors only someone familiar with the destination would catch. The information asymmetry that travel platforms held over travellers has not disappeared; it has moved.
What is changing, more durably, is the traveller’s relationship to planning itself. Itinerary-building used to feel like research. It now often starts as conversation — describing what you want, not searching for it. That conversational entry point rewards travellers who ask better questions and who know to verify what the AI produces. It rewards independent thinking over trusting that the platform has surfaced the best option.
Whether the travel industry builds tools that earn a place in that conversation — or watches from the side as travellers plan with tools they didn’t make, for needs they failed to anticipate — will shape the sector’s next decade more than any single booking trend or destination fad.
