After a deliberate approach, Airbnb is now planning to embed large language model-powered features into its app to assist users with listing searches, trip planning, and help hosts manage their properties, CEO Brian Chesky revealed on Friday.
During the company’s fourth-quarter conference call, Chesky stated Airbnb’s intention to expand its application of large language models for customer discovery, support, and engineering.
“We are developing an AI-native experience where the app doesn’t just search for you; it comprehends you. It will empower guests to plan their entire trip, assist hosts in optimizing their businesses, and enable the company to operate more efficiently at scale,” he explained.
Separately, the company announced it is testing a new feature that allows users to search for and ask questions about properties and locations using natural language queries.
Currently, Airbnb provides an LLM-powered customer service bot for some personalization and communication. The new AI search functionality is anticipated to “evolve into a more comprehensive and intuitive search experience that extends throughout the trip.”
When questioned by an analyst about the potential introduction of sponsored property slots within AI search, Chesky affirmed that the company’s priority is to first perfect the design and user experience.
“AI search is currently live for a very limited segment of our traffic. We are conducting extensive experimentation. Over time, we will be experimenting with making AI search more conversational, integrating it beyond just the trip search, and eventually, we will consider sponsored listings as a result of that,” Chesky elaborated, adding that Airbnb would explore designing an ad unit that aligns with the conversational search flow.
Chesky noted that Airbnb intends to leverage the AI expertise of its new CTO, Ahmad Al-Dahle (who previously worked on Meta’s Llama models), to utilize its vast amount of identity and review data to enhance the app’s usefulness.
Airbnb reported that its AI-powered customer support bot, launched in North America last year, now resolves one-third of customer issues without human intervention. Chesky also mentioned plans to enable customers to call the AI bot for support and to broaden language coverage for customer support.
“A year from now, if we succeed, significantly more than 30% of tickets will be handled by an AI customer service agent, in many more languages—all the languages where we have live agents. AI customer service will encompass not only chat but also voice,” he stated.
The company is also focusing on increasing internal AI utilization. Airbnb indicated that 80% of its engineers currently use AI tools, with a goal of reaching 100%.
Airbnb announced better-than-expected revenue of $2.78 billion for the fourth quarter, representing a 12% increase from the previous year.