Airbnb says it has harnessed artificial intelligence to accelerate product development, with company executives attributing major gains in speed and productivity to the technology. CEO Brian Chesky told reporters the firm has seen substantial improvements across engineering, search, and customer support as it integrates AI into both customer-facing features and internal tooling.
Concrete gains in speed and output
According to Airbnb, AI now produces a majority of the company's code, and the firm has reduced the time from concept to shipment by as much as 60 per cent on key initiatives. Management also says the number of features and improvements shipped in the most recent six-month period rose by nearly 80 per cent compared with the same span the previous year.
“Across some of our key initiatives, we’ve reduced the time from concept to launch by as much as 60%,” Brian Chesky said.
Those figures underscore a shift from proof-of-concept experimentation toward deep operational use of AI within one of the world's largest travel marketplaces. While Airbnb has previously rolled out incremental consumer features—such as review summaries and listing highlights—the company says it is increasingly embedding AI into search and transactional flows.
Where AI is being used
- Code generation: More than 50 per cent of Airbnb's codebase is now written with AI assistance, according to the company.
- Product development: Time from concept to shipment has been reduced by up to 60 per cent on select initiatives; feature output increased roughly 80 per cent year over year for a comparable six-month window.
- Customer support: AI agents handle a material portion of first-contact issues—nearly 45 per cent of customer problems initially reported to AI agents are resolved without human intervention.
- Search and onboarding: Airbnb is experimenting with agentic AI to improve search and host onboarding, while offering users a toggle to retain traditional search behaviour.
The company says the AI support bot launched in North America in 2025 and has since been expanded to more than 50 languages, with plans to integrate voice-call functionality.
Operational trade-offs and user choice
Airbnb has been cautious about forcing AI changes on customers. The company plans to introduce an AI-powered search option that allows natural-language queries but will maintain the existing keyword-and-filter search for users who prefer it. That approach recognises both the power and the friction of AI-driven interfaces—allowing a gradual, opt-in adoption rather than a hard migration for millions of users.
On the operational side, embedding AI has the potential to change staffing models, reduce cycle times and lower the marginal cost of product experimentation. The company’s expansion of multilingual AI support and moves into voice integration also point to broader ambitions to automate routine interactions at scale.
| Metric | Figure |
|---|---|
| Share of code written by AI | More than 50% |
| Time reduction from concept to launch | Up to 60% |
| Increase in features shipped (six-month comparison) | Nearly 80% |
| Customer issues completed by AI agents without human help | Nearly 45% |
For investors and industry observers, Airbnb’s metrics are a real-world data point for how AI can reshape software companies: speeding delivery, increasing feature velocity and automating large swathes of customer interaction. The company’s decision to preserve user choice for search functionality also signals a recognition that rapid technical progress must be balanced against user experience and trust.
As Airbnb scales these systems, regulators, hosts and guests will watch closely for impacts on job roles, platform fairness and data governance. For now, the company’s numbers offer a clear signal: AI is moving from experimental to foundational at one of the world’s leading consumer marketplaces.
