Personalizing travel at scale with OpenAI
OpenAI
Booking.com plugged OpenAI's models into its systems to make search and support smarter.
Based on reporting by OpenAI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Booking.com is one of those companies that has been quietly running machine learning for years, but its latest move is a shift in scale rather than kind. The travel giant has wired OpenAI's large language models directly into its own data systems, using them to power search, customer support, and what it calls intent-driven trip planning.
The pitch is simple enough: instead of a traveler typing rigid keywords into a search box and sifting through hundreds of listings, the system tries to understand what someone actually wants. A vague request like a quiet beach town with good food nearby can now get parsed into something usable, matched against Booking.com's own inventory of hotels, flights, and experiences, and returned as a shortlist that fits the vibe rather than just the filters.
Customer support is the other big piece. Travel questions are messy and time-sensitive. People ask about refund policies at 2 a.m. or need a booking changed twenty minutes before a flight. By connecting OpenAI's models to its internal knowledge and account data, Booking.com is aiming to resolve more of these moments instantly, without a human agent typing from a script.
What makes this different from a chatbot bolted onto a website is the integration with proprietary data. OpenAI's models are general-purpose reasoning engines; Booking.com's value is in decades of accumulated inventory, pricing, and behavioral data. Combining the two is the actual product here, not the novelty of a language model that can chat.
This fits a pattern OpenAI has been pushing hard in 2024 and 2025: selling itself less as a chatbot company and more as infrastructure for other businesses to personalize their own products at scale. Booking.com is a marquee customer for that story, and probably not the last big consumer platform to make this kind of bet.
My take — AI-written commentary, not fact-checked reporting
This is OpenAI's enterprise playbook working exactly as intended: get inside a company's data pipes so deeply that ripping the model back out becomes a multi-year project, not a vendor swap. I don't think travelers asked for an AI concierge, but I do think Booking.com asked for a way to cut support costs, and that's the real story every time one of these integrations gets announced with warm language about personalization.
Read more about this at: OpenAI