The Travel Industry Wants AI Agents. Airbnb Wants Control

When Airbnb CEO Brian Chesky discusses AI during the company’s Q1 2026 earnings call, he avoids framing chatbots as an inevitable interface for travel booking.

Chesky says current chatbot design “does not work for travel or e-commerce” because travel discovery depends on visual browsing, comparison, maps, and group decision-making rather than text conversations. Chesky identifies four structural problems with conversational interfaces: excessive text, limited direct manipulation, poor comparison across listings, and the “multiplayer” nature of travel planning.

Over the past year, companies including Google, Booking.com, and Expedia Group have introduced AI-assisted planning tools, itinerary builders, or conversational search products. At the same time, Airbnb expands its use of AI internally. Chesky says AI now writes nearly 60% of Airbnb’s new code, while more than 40% of customer support requests are resolved through AI systems without human escalation, according to TechCrunch and Airbnb’s shareholder materials.

Travel companies increasingly use AI for operational efficiency, customer support, and search assistance, while remaining cautious about allowing AI systems to control the booking transaction itself.

The Limits of Conversational Booking

Chesky’s criticism aligns with broader industry research around trust and reliability in AI-assisted travel booking.

An April 2026 study from Expedia Group finds that 68% of travelers still prefer booking through a trusted travel brand rather than directly through an AI platform. The survey, which includes more than 5,700 respondents across the United States, United Kingdom, and India, also finds that 66% of travelers would not trust AI to complete purchases or bookings on their behalf.

The same study finds travelers are significantly more comfortable using AI for lower-risk activities. About 53% are comfortable receiving AI-generated recommendations, 42% accept AI-based price monitoring, and 40% support AI itinerary planning, according to PhocusWire and Hospitality Net. Travelers show lower trust in AI systems for transactions involving payments, cancellations, or booking changes.

Xavi Amatriain writes in Expedia’s report that travelers “don’t have a technology problem with AI” but “a trust problem.”

Reliability concerns also persist in travel-focused AI systems. A 2025 study published on arXiv tests large language models on complex travel pricing scenarios involving bundled discounts and rule-based calculations. Researchers find that model accuracy drops sharply as pricing conditions become more complex, concluding that current LLM systems remain unreliable for “revenue-critical applications” without additional safeguards.

In March 2026, Reuters reported that OpenAI scaled back parts of its direct checkout ambitions after users showed stronger engagement with product discovery than transaction completion. Analysts at TD Cowen describe the move as a “stunning admission” that AI platforms replacing traditional applications may be delayed or structurally difficult in commerce-heavy industries like travel.

Even Google’s approach remains partnership-driven rather than fully disintermediated. In a November 2025 product announcement, Julie Farago writes that Google is building AI booking tools in collaboration with partners including Booking Holdings, Expedia Group, Marriott International, and IHG Hotels and Resorts. Google also states it has “no intention” of becoming an online travel agency itself. Users still complete bookings through travel partners rather than directly through Google’s AI systems.

Airbnb’s Broader Platform Strategy

While Airbnb avoids conversational booking products, it continues expanding the number of services inside its platform.

In May 2026, the company introduced hotel inventory in select cities, expanded local experiences, added food-related experiences, introduced luggage storage partnerships, and continues developing transportation-related offerings.

At the same time, Airbnb’s Reserve Now Pay Later product accounts for roughly 20% of global gross booking value during Q1 2026. The company also says 63% of nights booked now occur through Airbnb’s mobile app, up from 58% a year earlier, according to Airbnb shareholder materials.

Higher app usage gives Airbnb a larger share of direct customer traffic.

Airbnb continues expanding services inside its own platform while using AI primarily in operational workflows and avoiding conversational booking products in its consumer interface. The company has expanded into hotels, experiences, transportation-related services, and payment products inside the Airbnb app.