By Pablo Revale
March 10, 2026 · 6 Minutes
Adapted from Originally published in Unstoppable Hospitality on LinkedIn.
When Business Insider reported that OpenAI was pulling back from direct in-chat bookings, travel stocks surged. The market exhaled. The narrative hardened: AI is a discovery layer, not a booking layer. Trust lives with incumbents. LLMs are the new top-of-funnel. But read that move differently for a moment. OpenAI didn’t abandon transactions because it couldn’t do them. It stepped back because it’s building something more valuable first: unconditional user trust. Amazon spent years as a bookstore before it became the everything store. The sequence matters. Stepping away from in-chat bookings is not a matter of fear. It’s a strategic retreat.
Here is the tension no one wants to say out loud. LLMs currently send users to Booking.com and Expedia and collect a low discovery fee. The OTA closes the transaction, captures the customer email, owns the repeat booking, and earns the full margin. The LLM did the hard part and got the smallest check. That is not a sustainable business model for a company valued at over $300 billion. At some point, the board asks: why are we driving $40 billion in gross travel bookings and capturing 0.2% of it? The difference between a discovery fee and a transactional fee is not philosophical. It is a 200x difference in revenue per conversion. The incentive to close the loop inside the platform is existential. Why didn’t Google build an OTA? Back then, it required hundreds of millions in infrastructure, inventory, and trust-building. Today, an LLM with a payment toggle and 500 million users does it for a fraction of that cost. The barrier wasn’t will. It was tech barriers, economics, and a monopolized market.
The objections were identical. "Hotels will never allow a third party to own the booking relationship." "Travelers will never trust putting their credit card into a website." "The experience of choosing a hotel requires human judgment and personal touch." Every single one of those statements was spoken with confidence by smart, experienced leaders in the industry. Every single one of them was wrong, and that allowed the surge of new industry leaders that shaped the industry. By 2010, more than 50% of hotel bookings in Europe flowed through OTAs. The shift didn’t happen because the technology was perfect. It happened because one generation grew up treating it as normal and their normal became everyone’s normal.
63% of Gen Z uses AI tools daily for research and decisions. The average age of first regular ChatGPT users, 18 to 22, is now entering the workforce. 33% of Gen Z use LLMs to talk about sensitive topics. This is the part of the conversation the industry consistently misses, because it requires thinking about customers who are currently 16 years old. Today’s high schoolers didn’t adopt ChatGPT as a productivity tool. They adopted it as a companion. They ask it about their relationship problems. They process their career anxiety, their family conflicts, their financial stress. Now ask yourself: if you already told this thing about your breakup, illness, and financials, will you hesitate to add a credit card? A credit card is just data. And that generation learned that data is just trust made digital, from Airbnb, from Uber, from every app they used before they turned 18. Trust isn’t built by a brand. It’s inherited by a generation.
The unlock is simpler than people imagine. It’s not a technical breakthrough. It’s a UX moment. The day an LLM shows a prompt that reads "Save your travel preferences and payment details for seamless future bookings. You control what’s stored, anytime", millions of users will tap yes before they finish reading it. We have seen this before. The moment Uber asked "save this card for future rides?" changed consumer behavior permanently. The moment Apple Pay appeared in apps, card entry felt archaic within eighteen months. LLMs know this. They are not building toward discovery. They are building toward the moment the toggle is trusted enough to flip. There is a pattern worth noticing. The people who were most confident that paradigms wouldn’t shift were always the ones most invested in the current paradigm. The capital flowing into LLM infrastructure right now is not patient capital. It is impatient capital. It needs returns. And it has identified travel, a $10 trillion global industry, as one of the clearest paths to margin. What is your team doing today to prepare for potential scenarios?
