OpenAI just saved Booking.com

The piece argues that while the original thesis — that AI would weaken Booking.com’s control over travel discovery and booking — was broadly correct, the expected speed of disruption was overstated. OpenAI’s decision to scale back direct booking inside ChatGPT means AI platforms are, for now, staying in the discovery and referral layer rather than taking over the full transaction process. That shift benefits Booking.com, because the hardest and most defensible part of the business is not recommendation but the operational infrastructure behind payments, inventory, cancellations, merchant agreements, and global booking execution. Still, the threat has only been delayed, not removed: if AI platforms eventually gain transaction capabilities, Booking.com’s power could erode, but in the meantime it is well positioned to turn AI into just another referral channel it can outspend and optimise better than most competitors.

3 key takeaways:

1. OpenAI’s retreat strengthens Booking.com in the near term. Without in-chat transactions, AI platforms still need someone else to complete the booking, and Booking.com remains one of the strongest players to do that.

2. Booking.com’s real moat is operational, not just digital visibility. Its payments, inventory systems, merchant model, and global infrastructure are difficult and expensive to replicate, which is exactly why AI platforms have not displaced it yet.

3. Hotels should not mistake this for a long-term win. AI may still reshape distribution later, but for now the delay mainly gives Booking.com more time to dominate another referral channel, unless hotels invest in stronger direct-booking capabilities themselves.