AI now writes 60% of Airbnb's code

• Airbnb engineers now use AI to produce 60% of all company code
• AI support bots now resolve 40% of customer issues without humans
• AI tools help engineers manage tasks previously requiring 20 staff

Airbnb's Q1 2026 earnings reveal a dramatic AI integration, with the company's engineers now producing 60% of code with AI assistance. This technological leap is also enhancing customer support, reshaping how businesses operate.

AI Drives Code Production

During the first quarter of 2026, Airbnb experienced a monumental shift in its software development processes, largely driven by artificial intelligence. CEO Brian Chesky disclosed that AI tools were instrumental in generating approximately 60% of the code produced by the company's engineering teams. This marks a significant acceleration in their development cycles, placing Airbnb among industry leaders like Google, Microsoft, and Spotify who are similarly harnessing AI to expedite programming and software creation. Chesky highlighted the particular utility of AI in developing tools for Airbnb's API partners. These partners, who manage properties using diverse software systems, expressed a need for enhanced tools to improve their hosting capabilities. AI has provided this crucial leverage, enabling a single engineer to now oversee automated agents that can accomplish substantial work, a task previously requiring a team of around 20 engineers. This adoption of AI has empowered Airbnb to build more software for its API partners, thereby accelerating projects that were previously constrained by resource limitations.

AI Enhances Support

Beyond coding, Airbnb has significantly expanded its use of AI within customer support over the past year. Brian Chesky announced that the company's AI-powered support bot is now proficient in resolving 40% of customer issues without requiring human intervention, a notable increase from the approximately 33% handled earlier in the year. While AI's role in coding and support is advancing rapidly, Chesky acknowledged that the travel industry, including e-commerce, still faces hurdles in fully optimizing chatbot-based AI tools. He pointed out several limitations inherent in current chatbot designs for travel and shopping platforms. These include an overabundance of text, inadequate comparative features for different options, and the complex nature of bookings, which often involve multiple individuals, whereas chatbots are primarily designed for single-user interactions.