Truth that Matters. Stories that Impact

Truth that Matters. Stories that Impact

Technology

Airbnb Accelerates Product Development and Tests Natural Language Search with AI

Airbnb is accelerating its product development cycle using artificial intelligence, reporting significant reductions in development time alongside plans to test an optional natural language search experience for consumers.

What Happened

During Airbnb’s second-quarter earnings call, co-founder and Chief Executive Officer Brian Chesky shared that artificial intelligence has fundamentally altered how the company designs and ships features. Earlier this year, the company revealed that AI writes 60% of its code. Chesky stated that development time from initial concept to launch has dropped by as much as 60% across key projects, enabling an 80% increase in shipped improvements compared to the same six-month timeframe last year.

Alongside internal engineering gains, Airbnb is preparing to test an AI-driven search interface. Rather than replacing the existing search and filter tools, the company is introducing a toggle option. This allows travelers to query listings using conversational, natural language prompts that return visually rich results with AI-generated titles and real-time personalized listing highlights.

Key Highlights

  • Engineering Speed: AI tools write approximately 60% of Airbnb’s code, cutting concept-to-launch time by up to 60%.
  • Higher Output: The volume of features and improvements deployed increased by nearly 80% year-over-year during the first half of the year.
  • Targeted Areas: Internal AI integration spans search, user sign-up, checkout, payment systems, and simplified host onboarding.
  • AI Search Testing: A new toggle allows users to test conversational search with visual results and customized listing highlights without losing classic search filters.
  • Support Automation: An AI customer service bot expanded to over 50 languages resolves nearly 45% of initiated inquiries without human agents.
  • Financial Metrics: Second-quarter revenue rose 17% year-over-year to $3.6 billion, adjusted EBITDA climbed 21% to $1.3 billion, and support cost per booking dropped 16%.

Why This Matters

The operational metrics demonstrate how generative AI can directly affect software engineering productivity and operational costs at scale. For Airbnb, deploying AI in customer support has cut support costs per booking by 16% year-over-year, as nearly 45% of support interactions handled by AI are resolved without human escalation. On the consumer side, Airbnb is taking a measured approach by offering AI search as an opt-in toggle rather than forcing a conversational interface onto all travelers.

What to Watch Next

Airbnb plans to expand its AI customer support agent to handle voice calls later this year. The company will also evaluate user response to its conversational, visual AI search test as it rolls out to consumers.

Frequently Asked Questions

How is AI changing the search experience on Airbnb?

Airbnb is testing a toggle that allows users to switch to an AI search mode. In this mode, users can input natural language queries to receive conversational, visually focused search results alongside personalized, real-time listing highlights.

What percentage of Airbnb’s code is written by AI?

According to Airbnb, artificial intelligence currently writes 60% of the company’s code.

How has AI affected Airbnb’s customer service operations?

Airbnb’s AI support agent operates in more than 50 languages and resolves almost 45% of the customer issues it starts without human intervention, contributing to a 16% year-over-year reduction in support costs per booking.

Source: TechCrunch