Booking.com, Expedia & ChatGPT: What Hotels Need to Know.

In October 2025, the dynamics of hotel AI discovery changed in a structural way. Booking.com and Expedia both completed integrations with ChatGPT, embedding their inventory directly inside the world's most widely used AI assistant. For hotels listed on these platforms, this created an automatic AI distribution channel. For independent hotels without direct booking engine integration, it created something else: a mechanism through which AI-generated hotel recommendations route directly to OTA commission pages rather than the hotel's own website.

The Booking.com and Expedia ChatGPT integration means that when ChatGPT recommends a hotel, its default booking pathway — for properties that have not established direct booking integration — is to an OTA page. LuxDirect's London scans quantified the leakage rate: for independent luxury hotels without direct pricing synchronisation and API integration with Google Hotels, 73% to 93% of conversational AI-referred guests route to OTA pages. The integration that was supposed to expand hotel AI visibility has simultaneously deepened OTA dependency for hotels not prepared for it.

What exactly happened when Booking.com and Expedia integrated with ChatGPT?

In October 2025, both Booking.com and Expedia completed plugin and API integrations with OpenAI's ChatGPT platform. The integration enables ChatGPT to query real-time hotel availability, pricing, and inventory from these OTA databases when a user asks a travel planning question. When the AI recommends a hotel, it can surface live availability and provide a booking link — and for properties listed on Booking.com or Expedia, that booking link goes directly to the OTA's listing page for the property.

The Hotels Network also launched a direct booking application within OpenAI's ChatGPT app directory in 2026, powered by Lighthouse's Connect AI platform. This application enables properties to display verified, hotel-provided room photography, live rates, and custom descriptors to ChatGPT's user base — and provides a direct booking engine link rather than an OTA link. This represents the first direct booking channel inside ChatGPT, and it marks the beginning of a bifurcation in AI hotel distribution: properties with direct booking integration capturing guests directly, and properties without it paying OTA commission on AI-driven bookings.

How does the OTA–ChatGPT integration specifically affect independent hotels?

For independent hotels, the integration creates a structural disadvantage that compounds with AI visibility growth. As more travellers use ChatGPT for travel planning and booking, and as ChatGPT's default booking pathway for most properties routes to OTA pages, independent hotels face a scenario where improved AI visibility directly increases OTA commission costs. A hotel that works hard to appear in ChatGPT recommendations, but has not established direct booking integration, is growing its OTA commission bill — not its direct booking rate.

The 73–93% OTA leakage rate documented by LuxDirect represents the range across independent properties without direct booking API integration. Hotels at the higher end of that range — closer to 93% — are those with no Google Hotels connection and no direct booking engine that AI interfaces can access. Hotels at the lower end — closer to 73% — typically have partial Google Hotels connectivity but missing the real-time pricing sync that would allow AI to present direct booking options. No hotel without full direct booking API integration escapes significant OTA leakage in the current AI referral environment.

What does the 73–93% OTA leakage rate actually mean in revenue terms?

OTA commission rates for independent hotels typically run 15–20% of booking value. If a hotel generates $100,000 in AI-referred bookings annually and 80% of those bookings route to OTA pages rather than the direct engine, the hotel is paying $12,000–16,000 per year in OTA commission on bookings that originated from AI recommendations — a discovery channel the hotel invested in building. As AI-driven travel planning grows (generative AI referral traffic increased 357% year-over-year per Phocuswright), the absolute revenue lost to this leakage compounds proportionally.

The calculation changes entirely for hotels with direct booking integration. ChatGPT hotel referrals convert at 11.4% according to Similarweb data audits — more than twice the 5.3% rate for organic search. A hotel capturing AI referrals through its direct booking engine at 11.4% conversion, rather than paying 15–18% OTA commission on those same bookings, captures the full booking value plus the conversion premium. The revenue gap between integrated and non-integrated hotels will widen as AI travel planning volumes continue to grow.

Are OTA-listed hotels automatically AI-visible through the ChatGPT integration?

Not automatically — but OTA listing creates a significant AI visibility advantage. The ChatGPT integration with Booking.com and Expedia means ChatGPT can query these platforms' inventory in real time, but it does not guarantee inclusion for every listed property. ChatGPT still runs its own recommendation algorithm and will prioritise hotels that match the traveller's query attributes most specifically. A poorly described OTA listing for a hotel in the wrong category, with inconsistent room names, and no recent reviews is not guaranteed AI visibility just because it is on Booking.com.

What OTA listing provides is structural data access — ChatGPT can retrieve accurate pricing, availability, and property information from the OTA database in real time. This removes the data quality barrier for OTA-listed hotels at the query stage. But recommendation selection still reflects the property's positioning, description quality, review sentiment, and attribute match for the specific traveller query. Hotels with strong, well-maintained OTA listings will benefit more from the integration than those with minimal or inconsistent OTA profiles.

What can independent hotels do to compete with OTA AI visibility right now?

The first priority is Google Hotels direct booking integration — establishing real-time pricing and availability sync between your booking engine and Google's infrastructure. This is the most critical intervention for reducing OTA leakage because Google's AI interfaces (including Google AI Overviews and Google Gemini) prioritise direct booking pathways when they are available and verified. Without this connection, Google's generative search interfaces default booking routing to OTA partners.

The second priority is OTA listing quality — ensuring your descriptions, room names, amenity lists, and photography on Booking.com and Expedia are current, accurate, and consistent with your own website. Since ChatGPT now queries these platforms directly for inventory data, the quality of your OTA profile directly affects the quality of AI-generated descriptions of your property. Hotels that maintain excellent OTA profiles gain AI visibility advantages through the integration; those with stale or minimal profiles do not. The OTA listing is now both a direct booking channel and an AI data feed — neglecting it costs double.

Frequently Asked Questions

Sources: LuxDirect London AI Visibility Scans 2026 (OTA leakage rates); Similarweb Data Audits (ChatGPT conversion data); Phocuswright Travel Studies 2025–2026; The Hotels Network (ChatGPT direct booking app announcement 2026); MCP open-source release November 2024. RevPARGenius is an independent hotel market intelligence platform.