How Airbnb CEO Views AI Growth Without Heavy CapEx Spending

Airbnb's fourth-quarter results have delivered a compelling case study in how AI investments can translate directly into revenue acceleration

The company posted US$2.8bn in revenue for the fourth quarter of 2025, representing 12% year on year growth and surpassing the upper end of management guidance, while gross booking value climbed 16% to US$20bn, marking the strongest growth quarter in more than two years.

The holiday rental platform has managed to drive operational improvements through machine learning, while avoiding the capital-intensive infrastructure buildout that has concerned shareholders in other technology companies.

CEO Brian Chesky tells investors during the earnings call on 12 February 2025 that the acceleration "didn't happen by accident" but rather resulted from "a deliberate path we've been on for the past few years".

The financial implications extend beyond top-line growth.

Airbnb generated US$4.6bn in free cash flow throughout the 2025 fiscal year, whilst maintaining a 28% adjusted EBITDA margin in the fourth quarter.

This performance suggests the company has found a route to enhance customer experience and conversion without proportionally increasing costs, a balance that has proven elusive for many peers deploying AI technologies.

Product innovation translates to margin expansion

Management attributes approximately 200 basis points of nights growth and 300 basis points of gross booking value growth directly to recent product changes, many powered by AI.

The company has deployed what Brian described as "a custom AI agent trained on millions of our support interactions" which now handles a third of support issues without human intervention whilst reducing resolution times.

The cost implications could prove significant. In an interview with CNBC on 13 February 2025, Brian emphasises that "our investment in AI will not affect the P&L. I don't think you'll see it in the P&L. We do not have the huge CapEx cost base".

Rather than constructing proprietary foundation models, Airbnb is layering its booking data and customer interaction history onto existing large language models.

"We're not trying to build a foundation model," Brian explains. "We're going to leverage the best models and fine-tune them on our data."

The strategy also appears to be driving conversion improvements, with Brian noting on the earnings call that "traffic that comes from chatbots converts at a higher rate than traffic that comes from Google".

Operational efficiency meets revenue growth

Beyond customer service automation, Airbnb has focused on removing friction from the booking process, changes that carry both revenue and margin implications.

The introduction of transparent pricing addressing what Brian called "hidden fees, one of the biggest friction points in travel" appears to have reduced booking abandonment rates, though the company has not disclosed specific metrics.

The rollout of "Reserve Now, Pay Later" targeting higher-value properties has similarly contributed to booking acceleration.

"For the first time, guests in the US could book eligible stays paying US$0 upfront," Brian says, adding that "the response was immediate, driving booking acceleration in Q4, especially for larger, higher-priced homes".

This shift could improve both average transaction values and revenue per available listing, key metrics for the platform's unit economics.

From a financial strategy perspective, Brian has positioned AI not as experimental technology but as fundamental infrastructure. "From a business standpoint, I think AI is the best thing that ever happened to Airbnb," he tells CNBC.

"If you don't disrupt yourself, someone else will," Brian warns. "And we're not going to allow people to disrupt ourselves. We're going to disrupt ourselves first."

The fourth-quarter performance indicates this strategy may be working. With revenue growth accelerating, margins holding steady and free cash flow providing flexibility for continued investment, Airbnb has demonstrated that AI implementation can drive measurable financial returns without the capital intensity that has troubled investors in other technology stocks.

Whether this model proves sustainable as competition intensifies will likely determine the company's ability to maintain its current valuation premium.