Airbnb's public AI experiments have been cautious, but behind the scenes, the company is leaning hard on the technology to accelerate its entire product pipeline. That internal shift is now starting to surface in user-facing tools, with a new AI-powered search test on the horizon. During its second-quarter earnings call, CEO Brian Chesky revealed that AI is cutting the time from concept to launch by as much as 60% across key initiatives. He also noted that the number of features and improvements shipped in the first half of this year jumped nearly 80% compared with the same period last year. The company's internal code-writing AI, which generates 60% of its code, is a major driver of this efficiency. Main Developments Airbnb is finally testing an AI-driven search feature, a move that Chesky had previously resisted. He has argued that a plain chatbot interface is not the right fit for travel discovery, so the company has instead focused on integrating AI into search, discovery, and support. Read also: Disrupt 2026 Founders Guide: Building Enduring AI-Era Companies The new test adds a toggle that lets users switch from the traditional filters to a natural-language search. Results will appear in a visual, conversational format, with AI-generated titles and product descriptions that are personalized in real time. On the customer support front, Airbnb has deployed its AI agent in North America since 2025 and expanded it to more than 50 languages this year. Voice call support is planned for later in 2026. The company reports that nearly 45% of issues that start with the AI agent are resolved without human help, and support costs per booking have dropped 16% year-over-year. Background Airbnb's consumer-facing AI rollout has been deliberate, with features like review summaries and listing highlights appearing in isolation. Chesky has consistently cautioned against simply bolting a chatbot onto the platform, arguing that effective travel search requires more context than a typical chat interface provides. The company's internal adoption of AI has been more aggressive. Earlier this year, Airbnb said AI writes 60% of its code, and Chesky has credited AI with speeding up development in areas such as search, sign-up, checkout, and payments. Hosts have also benefited from quicker onboarding flows. Why It Matters Airbnb's approach offers a counterpoint to the industry-wide rush toward generative AI chatbots. By focusing on behind-the-scenes efficiency and carefully scoped consumer features, the company is testing whether AI can improve core travel experiences without disrupting user habits. The financial results for the quarter ending in June were strong, with revenue up 17% to $3.6 billion and adjusted EBITDA up 21% to $1.3 billion. That performance suggests Airbnb can invest in AI while maintaining growth, a balance that will be closely watched by competitors and investors. What's Next Airbnb will begin testing the AI search feature with a subset of users, though the company has not specified a timeline for a broader rollout. The success of the toggle approach will likely determine whether AI search becomes a permanent part of the platform. Voice support for the AI agent is slated for later this year, which could further reduce reliance on human customer service. As these features mature, Airbnb will need to balance personalization with user control, ensuring that AI enhancements feel like upgrades rather than impositions.