How Brian Chesky Uses an AI “Brain” to Run Airbnb

Brian Chesky’s case for AI at Airbnb starts inside the company. At Skift Global Forum in September 2026, the CEO described a system he built to answer a management problem: how to know what is happening across a large organization without spending so much time in meetings. He says it has cut his meetings in half. 2710

An internal AI system for executive visibility

Chesky called the system a “brain” or “intelligence corpus.” It draws on company information that includes 18,000 Keynote slides and 50,000 Google documents. Rather than relying solely on meetings to gather updates, he uses that internal information layer to get a broader view of Airbnb’s work. His claim of improved visibility is compelling as a description of how he manages, but it is not an independently verified measure of how completely the system represents the company. 2710

That internal focus distinguishes the management tool from a travel-planning chatbot. Airbnb has also put AI to work in customer support and software development; those uses offer a more concrete way to assess its broader ambitions. Reporting says Airbnb shipped more than 600 features in 2026, nearly twice the previous year’s number. Separately, its CTO said development teams were shipping roughly 80% more features than a year earlier. The figures indicate faster output, though they do not isolate how much of the increase AI caused. 26

The customer-support number needs particular care. One report says nearly 45% of issues that begin with Airbnb’s AI assistant are resolved without human intervention. That is not the same as saying AI resolves 45% of all customer-service requests. 6

How Airbnb’s claims compare with other CEOs’ AI plans

Chesky is describing changes he says are already visible in his own schedule and Airbnb’s operations. KPMG’s survey offers a different measure: executive expectations and spending. In its March 2026 release, KPMG reported that 77% of surveyed CEOs thought generative AI might have been overhyped over the past year but that its disruptive potential over the next five to ten years was likely underappreciated. Nearly 80% were spending at least 5% of their capital budgets on AI, according to that release. 36

KPMG materials provided here are not fully consistent: a separate 2026 data-highlights page puts the share allocating at least 5% of capital budgets to AI at 63%. The available excerpts do not explain the difference, so the spending figures should not be treated as interchangeable. Either way, a budget allocation measures commitment, not a realized productivity gain. 3336

The 2027 customer-facing test

At the forum, Chesky committed to launching an Airbnb AI agent in 2027 and described its planned interface as “much richer than a chatbot.” That puts a date on Airbnb’s longer-running ambition to make AI part of the traveler experience. It does not yet establish what the agent will be able to book, how well it will work or whether it will launch on schedule. 5626

The distinction is central to Chesky’s argument. His internal “brain,” reduced meeting time and reported delivery gains are claims about how Airbnb operates today. The 2027 agent is a promise about what travelers may get next. Success with the former may explain his confidence in the latter, but it cannot verify the future product in advance. 256