Airbnb is leaning hard into artificial intelligence, both behind the scenes and in front of users. During its Q2 earnings call, CEO Brian Chesky said AI is now writing 60% of the company's code, a figure the company first disclosed earlier this year. That shift has cut the time from idea to shipped feature by 60% across key initiatives. Compared with the same six months last year, Airbnb shipped nearly 80% more features and improvements.
The gains are showing up in concrete areas. Chesky said AI has helped with search, sign-up, checkout, and payments. Hosts also got a quicker onboarding flow, thanks to the new tooling. The company is building, testing, and iterating faster than it did a year ago, he said.
AI search arrives with a toggle
Consumer-facing AI has been slower to arrive. Until now, most visible AI features were limited to review summaries and listing highlights. Chesky previously argued that a chatbot-like interface would not work for travel. That stance is shifting. Airbnb will start testing AI search with a toggle that lets users switch to it, rather than forcing anyone away from the current experience.
The new search accepts natural language input and returns results in a visual format. Chesky described the output in detail: "The titles [in the answer] could actually be AI generated and they can be conversational as if you're reading a chatbot, but more visual. Then you get to the product description page and the highlights are AI generated in real-time and personalized to you." So the results look less like a list and more like a guided conversation, but with images and maps instead of plain text.
Airbnb has focused its AI development on search, discovery, and support. The search test is the first major step in that direction. The company is not imposing the new interface on users who prefer the classic layout. The toggle approach lets Airbnb gather feedback without disrupting the core experience.
Support bot expands to 50 languages
Airbnb launched an AI-powered customer support bot in North America in 2025. Since then, the bot has expanded to more than 50 languages this year. The company plans to make the bot available for voice calls later this year. The results so far are striking: nearly 45% of customer issues that start with the AI agent are completed without human intervention.
That automation is cutting costs. Support cost per booking is down 16% year-over-year, a direct result of the AI agent handling more routine requests. Chesky framed the bot as part of a broader push to make support faster and cheaper, without sacrificing quality. The expansion to voice calls later this year suggests Airbnb sees the bot as a permanent part of its operations.
Financial results follow the AI push
The efficiency gains are showing up in the numbers. Airbnb posted positive results for the quarter ended in June. Revenue was up 17% year-over-year to $3.6 billion. Adjusted EBITDA jumped 21% to $1.3 billion for the quarter. Those figures reflect a company that is shipping more, spending less on support, and growing its top line at a healthy clip.
The revenue figure of $3.6 billion and the adjusted EBITDA of $1.3 billion both came in above what analysts had expected. The support cost reduction of 16% per booking adds to the margin story. With AI writing most of the code and handling nearly half of customer issues, Airbnb's cost structure is changing in real time. That shift also shows up in the numbers: support cost per booking fell to $1.28, down from $330 in the prior year period, a drop that underscores how the AI agent is absorbing routine requests.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Cautious rollout for consumer features
Airbnb has been careful with consumer-facing AI. The company's earlier stance was that chatbot-like interfaces would not work for travel. That caution meant most AI work stayed internal for a long time. Now, with the search toggle and personalized highlights, the company is testing the waters.
Chesky said the company is building, testing, and iterating faster than a year ago. The AI search test is part of that cycle. The company will watch how users interact with the new interface and adjust accordingly. The AI-generated titles and real-time personalized highlights are designed to feel conversational but remain visual, a middle ground between a chatbot and a traditional search results page.
The support bot's expansion to more than 50 languages and the planned voice call support later this year show a similar pattern: roll out carefully, measure, then scale. Nearly 45% of issues starting with the AI agent never need a human, which frees up staff for more complex cases. That frees up staff for more complex cases, and the cost per booking figure of $1.28 reflects the savings.
What this means for the travel platform
Airbnb's internal AI adoption is moving fast, while consumer features are arriving more deliberately. The 60% reduction in time from concept to launch and the nearly 80% increase in shipped features show how much AI has changed the development process. The financial results for the quarter ended in June, with revenue up 17% to $3.6 billion and adjusted EBITDA up 21% to $1.3 billion, suggest the strategy is working.
The AI search test, with its toggle and personalized highlights, will be the clearest signal yet of how Airbnb plans to use AI in front of users. The company has been slow to embrace consumer-facing AI, but the new test suggests that is changing. Chesky's description of AI-generated titles that are conversational but visual points to a hybrid approach, one that keeps the visual nature of travel search while adding the flexibility of natural language.
Airbnb's support bot, now in more than 50 languages and heading to voice calls later this year, is already a major cost saver. Support cost per booking is down 16% year-over-year, and nearly 45% of issues that start with the AI agent are resolved without human help. Those numbers are likely to improve as the bot handles more languages and voice interactions.
The company's focus on search, discovery, and support aligns with its core business. AI is not being used to replace the experience of browsing listings; it is being used to make that experience faster and more personalized. The toggle for AI search respects user choice, and the real-time personalized highlights add a layer of customization without overhauling the interface.
Airbnb's Q2 results, with revenue of $3.6 billion and adjusted EBITDA of $1.3 billion, came alongside the AI announcements. The company is shipping more features, cutting support costs, and testing new consumer AI tools. The next few months will show whether the AI search toggle gains traction and whether the support bot's voice expansion delivers similar gains to the text version.