AI RACE— The AI Race
Business

Airbnb CEO Brian Chesky Calls for AI-Native Operating Systems and Rejects Chatbot Interfaces

Airbnb rolled out an AI-driven search experience alongside plans for voice agents, while CEO Brian Chesky argued consumer AI will stumble until the industry builds native operating systems and collaborative interfaces.

10/01/2026, 22:12
Business

Airbnb Debuts AI Search as CEO Outlines Agentic Future

Airbnb introduced an AI-powered search tool as part of its fall update, marking an expansion of artificial intelligence across the rental platform. In an interview with TechCrunch, Airbnb co-founder and CEO Brian Chesky stated that existing consumer AI has not yet matured, largely because standard chatbot interfaces fail e-commerce workflows and the tech industry lacks an operating system designed specifically around autonomous agents. Chesky revealed that Airbnb plans to deploy voice agents this fall for search and customer support, while researching multi-user "multiplayer" interfaces over the next three to six months to accommodate group travel planning.

Why Chatbots Fail E-Commerce and How Agents Must Interoperate

Chesky reiterated his longstanding critique of pure chatbot interfaces, arguing they serve single users poorly when shopping requires visual exploration and collaboration. Pointing out that travelers often enjoy browsing and anticipating vacations just as much as taking the trip, Chesky said conversational bots limit user choice to a handful of sequential responses. Instead, he anticipates a hybrid format that blends deterministic, designer-led controls with generative interfaces dynamically assembled by software—a space where startups such as Monogram and Wabi are also experimenting.

Looking forward, Chesky envisions Airbnb transforming into a constellation of specialized AI agents—spanning discovery, customer service, and other functions—coordinated under a macro Airbnb agent. This framework would allow the company to communicate across external platforms via open protocols such as Model Context Protocol (MCP). However, Chesky noted that existing consumer agents like Instinct and Muse struggle with complex bookings on both hotels and Airbnbs. To resolve this, he argued that agent developers need to provide deeper software development kits rather than attempting to reduce full-featured apps into bare data layers for a generic text box.

The Operating System Bottleneck and Big Tech Platforms

According to Chesky, consumer AI faces an architectural limitation: software today is built atop legacy environments like iOS, macOS, and Windows rather than an operating system operating at the kernel level for AI. Chesky recalled cautioning OpenAI CEO Sam Altman about this constraint when ChatGPT launched its app store without a foundational operating system and dedicated SDK, which he argued hindered functionality.

Chesky pushed back against two prevailing assumptions in Silicon Valley: that traditional user interfaces will entirely vanish in favor of voice and text prompts, and that software will become completely generative. Because professional designers generate superior software interactions compared to average users, Chesky expects apps to remain anchored to mobile ecosystems until tech giants build a replacement layer. He concluded that the transition from standalone apps to fully autonomous agents ultimately rests on platform holders like Apple and Google architecting genuine AI-native operating systems.

◗ Sources

TechCrunch10/01

Related stories