Airbnb CEO says AI travel agents currently have minimal impact on bookings

The current state of AI agents

Airbnb CEO Brian Chesky recently addressed the growing conversation around AI agents and their potential to disrupt the travel industry. During the company's third quarter earnings call, he noted that agents like Meta's Muse and Instinct AI are currently bringing in a very small amount of business. While these tools have gained attention in tech hubs, their actual impact on Airbnb's nearly half million daily bookings remains minimal.

Chesky observed that the traffic coming from these third-party agents actually peaked in early October and has not seen any growth since that point. He described the volume of bookings from these specific platforms as 'very non-material' for the company's overall operations. This suggests that while the technology is capable of making bookings, it has not yet reached a level of adoption that threatens the company's core business model.

The sentiment shared by Airbnb is echoed by other major players in the travel and delivery spaces. Executives from Booking.com and Expedia also reported during their recent earnings calls that AI agents have resulted in very little traffic so far. Even DoorDash CEO Tony Xu noted that third-party agents had generated only several hundred orders in total as of last week.

Why general AI fails at travel

The reason for this low adoption may lie in the fundamental difference between routine tasks and aspirational experiences. Chesky argues that current general-purpose AI agents are well-suited for simple, repetitive actions like scheduling an appointment or making a restaurant reservation. However, planning a vacation is a different psychological process that involves discovery and emotional investment.

According to Chesky, travel is inherently aspirational, meaning that the act of discovering where to go and where to stay is a core part of the enjoyment. He cited data showing that 80% of travelers actually enjoy the time they spend planning their trips. Because people want to look at photos, compare different homes, and make decisions with their travel companions, they may not want to offload that entire experience to a bot.

The company's data supports the idea that users are highly engaged in the manual search process. For instance, customers typically look at a dozen different listings and often swipe through 25 photos for every single listing they view. This level of interaction suggests that the 'discovery' phase is a primary way users interact with the brand, making it difficult for a simple text-based agent to replace the visual and social experience.

The shift toward specialized tools

Because general-purpose agents are not the best fit for complex travel needs, Airbnb is preparing to develop its own solution. Chesky announced that the company plans to begin rolling out its own specialized AI agent in the first half of next year. This tool will be built specifically for travel to provide guests with a more custom and tailored experience than a generic assistant could offer.

The strategy reflects a belief that people will increasingly turn to specialized agents rather than trying to use a single AI to do everything. By creating a tool that understands the nuances of travel—such as the importance of photos and host communication—Airbnb aims to capture the planning process more effectively. This move allows the company to maintain its role as the primary platform for discovery while still embracing the efficiency of AI.

Chesky expressed optimism about the technology, stating, "I don't see them as a threat." Instead, he views these agents as a potential opportunity for more people to discover Airbnb. He described them as a 'great handoff' to the platform, where an agent might help a user find a starting point before they transition into the deeper, more interactive booking process on the Airbnb site.

User behavior and work habits

The company also shared some interesting insights into when and how people are currently using their platform. A significant portion of Airbnb's traffic occurs during standard work hours, specifically between 9 a.m. and 5 p.m. Chesky noted that there is clear evidence that people are secretly using the app while they are supposed to be working, perhaps as a way to mentally escape their jobs.

n

This behavior underscores the high emotional engagement users have with the platform. When people are browsing listings during the day, they are often looking for an escape or a future goal, which reinforces the idea that the planning process is a leisure activity. Because it is a leisure activity, users are less likely to want to automate the entire experience away, preferring to keep the human element of choosing and dreaming involved.

Despite these nuances, Airbnb remains in a strong financial position as it moves toward these new AI integrations. The company reported its fastest revenue growth in three years during the last quarter, with a 19% year-over-year increase in revenue. This growth provides the company with the resources to invest in the specialized AI tools it plans to launch next year to stay ahead of the competition.