Airbnb is rapidly integrating artificial intelligence (AI) into its core offerings, as announced by CEO Brian Chesky during the fourth-quarter earnings call. The company aims to leverage large language models (LLMs) to enhance various aspects of its platform, including search functionalities, trip planning, support for hosts, and customer service, all contributing to a more “AI-native” user experience.
New AI features are designed to enable the platform not only to respond to user queries but also to “understand users,” providing personalized travel recommendations and comprehensive process support. This innovation will facilitate trip planning for guests, allow hosts to manage their listings more efficiently, and improve operational efficiency at scale.
Currently, Airbnb is testing a natural language search function with a limited user group, enabling conversational inquiries about listings, locations, and other travel-related queries. This feature is intended to create a more immersive and intuitive search experience throughout the customer journey. While the focus is on refining design and user experience, the company plans to explore new advertising formats aligned with conversational search logic in the future.
In customer service, Airbnb's AI chatbot, introduced in North America last year, manages approximately one-third of user requests without human intervention. The company aims to expand this system further, incorporating voice interactions and additional languages, with the goal of having over 30% of service requests handled by AI within a year.
Additionally, Airbnb recently appointed Ahmad Al-Dahle, a former Meta Llama model team member, as Chief Technology Officer to utilize his AI expertise, enhancing the system's understanding and recommendation capabilities. Internally, 80% of engineers currently employ AI tools, with the aim of achieving full coverage to boost engineering productivity and speed up product iterations.
