Airbnb Says AI Now Writes 60% of Its Code

Benzinga reported Thursday that Airbnb AI coding tools now generate 60% of all new code at the company. The disclosure came during the company’s first-quarter earnings call, where CEO Brian Chesky also announced an 18% revenue jump to $2.7 billion.

AI Agents Take the Wheel on Software Development

Chesky told investors that a single engineer equipped with AI can now deliver the output of a team of roughly 20. The agents operate with humans keeping oversight rather than running autonomously. Chesky framed the shift in direct productivity terms, saying AI tooling creates enough leverage to pursue software projects the company previously had no bandwidth to attempt. Airbnb carries a market capitalisation of approximately $84 billion, and shares closed near $141.50 on Friday before the call.

Also Read: Meta Says AI Agents Will Join Your Group Chats — Here Is What We Know

Customer Support Automation Also Gains Ground

The Airbnb AI coding story is only part of a broader automation push underway at the firm. Its AI-powered customer support agent now closes roughly 40% of incoming issues without escalating to a human representative. That figure marks a meaningful improvement from the 33% resolution rate the company recorded earlier in 2026. The progress signals that AI deployment is moving beyond engineering teams and into front-line operations at scale.

Big Tech’s AI Coding Race: Background

The race to hand software development to AI models is accelerating across the industry. Anthropic CEO Dario Amodei said at the World Economic Forum in January that some of his own engineers have stopped writing code by hand entirely. He predicted AI could handle the majority of software engineering tasks within a six-to-twelve-month window. Airbnb’s disclosure lands squarely within that broader industry narrative and adds hard company-level data to what has largely been executive-level prediction.

Also Read: Anthropic Raises $2.5 Billion as AI Arms Race Intensifies

Chesky Flags What AI Still Cannot Do for Travel

Chesky was not entirely bullish on AI’s current shape. He cautioned that no company has yet cracked AI-native travel or e-commerce experiences. Current chatbot interfaces carry too much text, he argued. They lack intuitive comparison features, offer no direct manipulation of results, and are built for solo users at a time when most travel bookings involve group decision-making. The comments suggest Airbnb sees the consumer-facing AI product gap as its next opportunity, not yet a solved problem.

Read Next: OpenAI Targets $3.4 Billion in Annual Revenue as ChatGPT Enterprise Grows