Airbnb, a short-term rental platform, said that its self-developed AI customer service system has currently handled about one-third of customer support issues in North America, and the company is preparing to bring this feature to the global market. If things go well, Airbnb expects that within a year, more than 30% of global customer service tickets will be handled by AI voice and chat systems, covering all language areas with human customer service.
In a fourth-quarter earnings call this week, CEO Brian Chesky said: "We think this will have a huge impact because not only will it reduce the cost base of Airbnb customer service, but the quality of service will also achieve a significant leap." This statement seems to imply that he believes that AI will do a better job than human customer service at resolving certain issues.
The company also highlighted the recently hired chief technology officer Ahmad Al-Dahle, who was poached from Meta for his AI expertise and will be responsible for driving Airbnb's efforts to create AI-native experiences. Chesky revealed that under the guidance of Al-Dahle, Airbnb is preparing to launch an application that can not only search for users, but also "get to know you." "It will help guests plan their entire trip, help hosts run their businesses better, and help companies operate more efficiently at scale," he explained. That's exactly why Airbnb brought in Al-Dahle.
Chesky emphasized that Al-Dahle is one of the world's top AI experts. He has worked at Apple for 16 years and recently led Meta's generative AI team that built the Llama model. "He has a knack for combining large-scale technology with world-class design, which is exactly what we need to transform the Airbnb experience," Chesky noted.
Similar to other businesses facing AI disruption, Airbnb's leadership is promoting the idea that the company has a unique database and product that no other AI chatbot can replicate. "Chatbots don't have our 200 million verified identities, our 500 million proprietary reviews, and they can't send messages to hosts — which 90 percent of guests do," Chesky told analysts on the earnings call. Instead, he floated the idea of layering an AI layer on top of the Airbnb experience, claiming it would help accelerate growth.
In terms of financial performance, the company forecasts revenue growth of "low double-digit" percentages this year. In the fourth quarter, Airbnb achieved revenue of $2.78 billion, exceeding expectations of $2.72 billion. For the current quarter, the company expects revenue in the range of $2.59 billion to $2.63 billion, above Wall Street forecasts of $2.53 billion.
Investors remain concerned about whether AI platforms pose risks in the long term, especially assuming they enter the short-term rental market. But Chesky pushed back, noting that Airbnb is not just a consumer-facing app but also includes a host app, customer service, and protections like insurance and user verification. "We spent 18 years building this, and the platform handles more than $100 billion in payments," he said.
Meanwhile, AI chatbots function like search engines, delivering top-of-funnel traffic. Chesky noted that these traffic also convert at a higher rate than traffic from Google, suggesting the shift to AI will benefit Airbnb.
The company is already leveraging AI-driven search functionality, which is currently enabled for a "very small percentage" of Airbnb traffic, and is experimenting with making searches more conversational. In the future, the company plans to integrate sponsored listings in search.
After Spotify told investors this week that its best developers hadn't written a line of code since December because of AI, Airbnb provided a broader metric of internal AI adoption. The company says that 80% of its engineers are currently using AI tools and is working to increase this proportion to 100% as soon as possible.