Airbnb Tests AI Search After Automating 45% of Support

Airbnb’s approach to AI is distinct from that of its rival Booking Holdings. Chesky chose to tackle customer service first, calling it the hardest problem in AI, before extending AI to the traveler-facing search and discovery experience. That order reflects a bet that owning the AI experience at every step—support, search, discovery, and safety—keeps travelers inside Airbnb’s ecosystem. In contrast, Booking Holdings’ CTO argues that the platform completing the transaction holds the real value and risk, even if an AI assistant plans the trip.

Internally, AI is reshaping how Airbnb builds its own product. The company says AI has cut the time from concept to feature launch by as much as 60% and now writes roughly 60% of its new code. This acceleration underpins features across sign-up, checkout, and payments, meaning travelers may experience smoother flows without realizing AI is working behind the scenes. Meanwhile, AI-powered safety tools like anti-party technology flag only about 0.15% of global bookings as potential risks.

As both companies converge on AI, they are starting from opposite assumptions about where its value lies. Airbnb is investing in a broad, integrated AI layer, while Booking Holdings focuses on the transactional core. The coming months, with Airbnb’s planned AI search and home comparison features, will show which approach resonates with travelers and hosts.