You’re reading AI at the Top - Our Sunday series, where we share how leading enterprises and their executives use AI in business.
Airbnb’s CEO Brian Chesky announced in August 2025 the enterprise is transforming into an “AI-first application.”
We saw this coming. Almost all non-tech companies are eventually investing in AI to serve their customers faster, efficiently.
In Q2 2025, Airbnb reported $3.1 billion in revenue, up 13% YOY. While numbers look decent, there’s always more to accomplish with AI.
In today's AI at the Top, we'll explore whether Airbnb’s AI-first approach is just a trending phrase or if the enterprise actually has strong use cases and business impact.
Enterprises with strong proprietary data are winning the AI game. Airbnb is the same.
Most travel platforms track transactional data like dates, prices, and bookings. Airbnb captures something more valuable: Behavioural data.
Think unstructured data of human interactions.
Every search query you make, every listing you click but don't book, reviews you write, messages between host and guest, all of it feeds Airbnb's AI.
The enterprise is capable of capturing behaviours because of its two-sided marketplace. Hosts and Guests.
The value of the Airbnb platform for any single user is enhanced by the cumulative data contributed by all previous users. 17 years of data compounding every second in real time.
The proprietary dataset can’t be replicated by new competitors, allowing Airbnb's AI to solve for CX, personalisation, and revenue optimisation.
What are Airbnb’s AI use cases?
Airbnb built a customer service agent to solve the hardest problem first.
“Customer service is the hardest problem because the stakes are high, you need to answer quickly, and the risk of hallucination is very high.”
Airbnb built a custom AI agents on 13 different models, trained on tens of thousands of actual customer conversations.
When you contact support, it taps into your booking history, current reservations, and past interactions to provide accurate, specific help.
The agent doesn’t just know how to cancel a reservation. It knows which reservation you want to cancel without asking. It understands context.
The company quietly rolled this out to half of the US customers in April 2025.
Result = 15% reduction in guests needing humans for support.
With 134 million nights and experiences booked in Q2 2025 alone, the 15% reduction translates to millions of interactions handled by AI instead of humans.
The company plans to expand this agent to more languages (currently English only) and eventually give it the ability to search, plan, and book trips.
🧠 Enjoying this read? Share the smartest part with your followers. Click to share.
Maximising the revenue for hosts by analysing the factors hosts don’t have time for
Hosts face a classic pricing dilemma. Highly priced bnbs go unbooked. If lower prices, revenue takes a hit.
Traditional pricing relied on static factors and couldn’t account for dynamic market changes. Hosts spent hours manually checking competitor rates and adjusting prices.
Airbnb's AI analyses multiple variables in real-time:
Location and season patterns
Local events like concerts, conferences, sports games, and holidays
Competitor rates in the area
Historical booking patterns for similar properties
Property’s features and amenities
On top of it, hosts can customise how aggressive the algorithm should be. Some prioritise higher occupancy rates. Others want to maximise revenue per booking.
The AI handles the long, boring analysis in real-time, allowing the hosts to focus on creating great experiences.
YOU DECIDE
🌟 Want to be featured in the next issue? Reach out with your best AI use case and we’ll spotlight it.
Solving trust and safety issues faced by unauthorised parties
Guests booking properties for unauthorised parties damage hosts and hurt Airbnb’s brand reputation.
Manual screening is of course impossible at the scale of millions.
The AI analyses risk patterns like one-night stays during public holidays, last-minute bookings from new accounts with no history, previous party patterns, etc.
The enterprise blocked 1.4 million unauthorised parties, and the host complaints have reduced by 50%.
Personalised Search and Recommendations
Airbnb's AI moved beyond basic filters like dates, price, and location.
The system now analyses click patterns, booking history, browsing behaviour, and host preferences to rank listings by the likelihood you’ll actually book.
If you keep clicking on properties with outdoor spaces and mountain views, those listings get prioritised, even if you never explicitly filtered for them. It understands intent.
This reduces search time for guests and increases booking conversion rates for hosts.
Where is Airbnb heading?
“I think you can't do travel planning without AI going forward.”
Airbnb is in the middle of that transformation into an AI-native app.
The company acquired GamePlanner AI, an AI startup focused on travel technology. This signals serious intent in building advanced AI capabilities in-house.
Brain’s vision is for Airbnb’s AI to book flights, reserve restaurants, suggest and book activities, handle itinerary changes, and manage your trip end-to-end.
Also, the enterprise is expanding beyond homes into experiences (guided tours, classes, activities) and services (photographers, chefs, wellness treatments).
Multiple agents specialise in functions like booking, customer service, trip planning, dispute resolution, etc. The agents will learn from each other, creating a holistic experience for its users.
This positions Airbnb to compete more aggressively with traditional travel agencies and hotels.
Why use multiple apps when one AI agent handles everything?
Sources:
Articles from Klover.ai, Digital Defynd, Constellation Research, Entrepreneur
Coverage from TechCrunch, Fortune, PYMNTS
YOUR TURN