Airbnb Says AI Is Slashing Feature Build Times as It Prepares a Test of Natural-Language Search

Airbnb says artificial intelligence is now speeding up the way it builds products, cutting the time from idea to launch by as much as 60% as the company prepares to test a new AI-powered search experience. The shift matters because it shows Airbnb is using AI first as an internal engineering accelerator, while moving more cautiously on customer-facing features.

During the company’s second-quarter earnings call on August 7, co-founder and chief executive Brian Chesky said AI is helping Airbnb ship more features, iterate more quickly and improve core travel flows such as search, sign-up, checkout and payments. He also confirmed that the company plans to begin testing AI search, though it will not replace the app’s traditional filters outright.

Airbnb’s approach reflects a broader pattern in consumer technology: many companies are rapidly adopting AI behind the scenes, but they are still testing how far users are willing to go with conversational interfaces in place of familiar product controls. In Airbnb’s case, the company is trying to preserve the current search experience while offering an opt-in alternative that can interpret natural-language requests and return visual results.

What Airbnb said about AI’s impact on product development

Airbnb says AI is not just a side experiment anymore; it has become a practical tool for building software faster and at larger scale. Chesky told investors that AI has materially changed the company’s product development cycle over the past year.

Chesky said the company is building, testing and refining products faster than it was a year ago, adding that some key efforts have seen the time from concept to launch cut by up to 60%. He also said Airbnb has increased the number of features and improvements shipped this year by nearly 80% compared with the same period six months earlier.

The company’s previous disclosure that AI was writing around 60% of Airbnb’s code now appears to be translating into measurable output. Rather than framing AI as a future productivity boost, Airbnb is presenting it as an existing lever that is already changing how teams work.

That matters for a company of Airbnb’s size because product velocity can influence everything from user retention to host satisfaction. Faster delivery means Airbnb can respond more quickly to market shifts, refine booking flows and test new product ideas with less delay.

Why does this matter for Airbnb’s business?

It matters because Airbnb’s core business depends on both trust and convenience. Travelers need search tools that are accurate and easy to use, while hosts need smooth onboarding, reliable payments and less friction in support. If AI can shorten the development cycle, Airbnb can improve those experiences more quickly and potentially lower operating costs at the same time.

The company said AI has been useful in several high-value areas, including search, registration, checkout and payment processing. Airbnb has also used the technology to streamline host-facing workflows, including faster onboarding.

How will Airbnb’s AI search work?

Airbnb’s first test of AI search will let users type questions or requests in natural language and receive results in a visual format, but only if they choose to switch it on. The company is adding a toggle so users can decide whether to use the traditional search-and-filter system or the new AI-driven option.

That opt-in design is important. Airbnb is not forcing a conversational interface on travelers who already know how to search by price, location, dates and amenities. Instead, it is trying to make AI available to people who want to describe what they are looking for in a more open-ended way.

According to Chesky, the idea is for the results to feel conversational but remain visual and grounded in travel listings. He said the titles in the answers could be generated by AI, the listing highlights could be produced in real time, and the output could be personalized to the individual user.

That approach suggests Airbnb is trying to blend the familiarity of search with some of the flexibility of a chatbot. The company appears to believe that travel discovery is a better fit for AI than a pure chatbot interface, because travelers often need visuals, structured comparisons and clear details before they book.

Why not just launch a chatbot?

Because Airbnb believes a generic chatbot is not enough for travel planning. Chesky has argued that conversational interfaces alone do not fully solve the needs of people searching for places to stay, where users usually want to see properties, compare attributes and make decisions based on images and facts as much as language.

That view helps explain why Airbnb’s consumer-facing AI rollout has been gradual. The company has preferred targeted features such as review summaries and listing highlights rather than a fully conversational front-end that would replace the app’s existing structure.

In practical terms, Airbnb is testing whether AI can make search more intuitive without undermining one of the app’s most established behaviors. For a travel platform, that balance is critical: too much novelty can confuse users, but too little ambition risks making the product feel dated.

Where Airbnb is already using AI behind the scenes

Airbnb’s most mature AI deployment is in customer support, where the company says automation is already handling a meaningful portion of the workload. The company launched an AI-powered support bot in North America in 2025 and has since expanded it to more than 50 languages.

Airbnb also said it plans to make the bot available for voice calls later this year, extending the system beyond text-based interactions. That would broaden the use case from routine help requests to a more flexible service layer capable of serving users in multiple formats.

How effective is the support bot?

Airbnb says nearly 45% of issues that begin with its AI agent are resolved without a human employee stepping in. The company also said support cost per booking has fallen 16% compared with the same period a year earlier.

Those figures suggest AI is doing more than answering simple questions. It is helping Airbnb absorb a substantial share of repetitive customer requests, which can lower service costs while freeing human agents to focus on more complex or sensitive cases.

• Launched in North America: 2025
• Expanded to more than 50 languages
• Voice-call support planned later in 2026
• Nearly 45% of issues resolved without human intervention
• Support cost per booking down 16% year-over-year

For Airbnb, support automation has an additional strategic benefit. Travel disruptions, cancellations and booking questions can create spikes in service demand. An AI layer can help manage those peaks more efficiently, especially in a business where volume can rise quickly around holidays, major events and peak travel seasons.

What does the company’s earnings report show?

Airbnb’s latest financial results suggest the company is pairing AI investment with solid operational performance. For the quarter ending in June, revenue rose 17% year-over-year to $3.6 billion, while adjusted EBITDA increased 21% to $1.3 billion.

Those numbers matter because they show Airbnb is not adopting AI from a position of weakness. Instead, the company is using a period of healthy growth to press for more efficiency and faster product delivery.

In other words, AI is being presented as both a competitive tool and a cost-management tool. Faster development can help Airbnb introduce new features, while improved support automation can help protect margins.

How does this compare with Airbnb’s earlier AI strategy?

The company’s current stance is more measured than the hype around many consumer AI products. Airbnb has been clear that it does not believe every interface should become chat-first. Instead, it is focusing on areas where AI can improve utility: discovery, support and internal engineering.

That is a notable contrast with companies that have rushed conversational AI into the main user experience before fully solving quality, accuracy and trust concerns. Airbnb is signaling that travel is a domain where context matters and where users may prefer hybrid designs that combine AI assistance with familiar controls.

Why is Airbnb being careful with consumer AI?

Airbnb is being careful because travel decisions are high-stakes and often personal. Booking a stay is not the same as asking a chatbot to draft an email or summarize a document. Users want confidence that the information they see is correct, current and visually understandable before they commit money to a reservation.

That helps explain why the company’s consumer AI features have been limited so far. Review summaries and listing highlights are useful, but they do not fundamentally alter how people search. The new AI search test is a larger step, yet Airbnb is still preserving the traditional interface as the default.

The company’s caution may also reflect a broader lesson in product design: AI can be powerful, but only if it fits the task. A booking platform built on trust, filters and photos may benefit more from AI that enhances search and personalization than from a universal chatbot that tries to do everything.

What users may notice first

For travelers, the earliest visible change may be a more natural way to describe what they want. Instead of selecting dozens of filters, they could ask for a family-friendly apartment near a specific neighborhood, or a quiet place suitable for remote work, and get results presented visually.

If the test expands, users may also see more personalized listing descriptions and more dynamic highlights generated in real time. That could make the shopping process feel more tailored without removing the structure people rely on today.

What this says about AI in consumer tech

Airbnb’s update is part of a wider shift in consumer software: companies are increasingly using AI to improve internal throughput before they fully redesign the customer experience. That sequence is becoming common because it is easier to prove value behind the scenes than to persuade users to adopt a new interface model.

For investors, Airbnb’s numbers are particularly telling. Faster shipping and lower support costs are exactly the kinds of operational gains that can strengthen margins. At the same time, a smarter search experience could make the product more competitive in a crowded travel market.

For the industry more broadly, Airbnb’s strategy highlights a pragmatic version of AI adoption. Rather than chasing headlines with a fully automated front end, the company is using AI where it can improve speed, reduce friction and support the core booking experience.

Timeline: Airbnb’s AI rollout so far

Airbnb’s AI strategy has evolved in stages, with most of the progress happening out of view before the current search test announcement.

Bottom line

Airbnb is using AI to accelerate the way it builds software, resolve more customer support issues automatically and prepare a cautious test of natural-language search. The company says those changes are already helping it move faster, reduce support costs and ship more product improvements while still growing revenue and profit.

The bigger story is not that Airbnb is embracing AI, but how it is doing so. The company is prioritizing practical gains over flashy reinvention, suggesting that the next wave of AI in travel may be less about replacing search and more about making the existing experience smarter, faster and more personal.