Airbnb CEO Brian Chesky has clarified the company’s approach to artificial intelligence, confirming that Airbnb does not rely on OpenAI’s latest models in production. Instead, the platform utilizes a multi-model strategy, integrating 13 different AI models—including Alibaba’s Qwen—to power its AI-driven customer service agent.
Diversified AI Strategy for Flexibility and Reliability
Chesky explained that this diversified approach enables Airbnb to optimize performance, tailor responses to different customer needs, and reduce dependency on any single AI provider. The AI agent handles inquiries, offers personalized recommendations, and assists hosts and guests efficiently across the platform.
Industry experts note that Airbnb’s strategy reflects a growing trend among tech companies seeking flexibility and reliability in AI deployments. By leveraging multiple models, companies can balance speed, cost, and accuracy while mitigating risks associated with over-reliance on one vendor.
Commitment to Responsible AI Use
The company emphasized its dedication to responsible AI usage, ensuring that customer interactions are transparent, accurate, and secure. Chesky’s comments come amid heightened scrutiny of how tech platforms deploy large language models in production, highlighting the importance of ethical and accountable AI practices.
Expanding AI Capabilities Across the Platform
Airbnb continues to enhance AI features across its ecosystem, aiming to improve user experience while maintaining trust and safety for its global community of hosts and guests. The multi-model approach positions the company to innovate responsibly while offering scalable and reliable AI-powered services.
