Airbnb CEO: AI Agents Need a Dedicated Operating System

Overview

Airbnb has introduced a new AI-powered search feature as part of its recent fall update, marking a significant pivot in how the hospitality giant approaches artificial intelligence. Co-founder and CEO Brian Chesky has long argued that a chatbot-only interface is insufficient for travel discovery, a stance that has guided the company’s cautious entry into the AI space.

During a recent interview, Chesky elaborated on why the current search tool is merely a stepping stone rather than the final product. He outlined a vision where AI agents require a dedicated operating system to function effectively, emphasizing the need for interfaces that support collaborative planning and adaptive user experiences.

Rethinking the Interface: Beyond the Chatbot

Chesky maintains that traditional chatbots are ill-suited for e-commerce and browsing tasks. He argues that conversational interfaces force users to take multiple turns to reach a desired result, creating friction in the shopping journey. However, he also noted that the current Airbnb search interface is not the endgame for travel technology.

"I think what you’re seeing today is also not the endgame for e-commerce or for travel or shopping. I think there’s going to be some interface... something between a chatbot and what you see in the first version we shipped."

The CEO highlighted a key psychological insight: people often derive more pleasure from planning and anticipating a trip than from the travel itself. Consequently, Airbnb aims to preserve the "library experience" of dreaming and browsing, rather than reducing travel discovery to a simple transactional command. This philosophy drives the company’s exploration of new interface models that balance speed with inspiration.

The Rise of Multiplayer AI

A critical limitation of existing AI tools is their design for single-user interactions. Chesky pointed out that travel planning is frequently a collaborative activity involving family members or friends. To address this, Airbnb is focusing on developing "multiplayer" AI interfaces over the next three to six months.

This shift represents a move away from solitary assistant models toward shared digital spaces. By enabling multiple users to interact with AI simultaneously, Airbnb hopes to facilitate group decision-making without fragmenting the experience. This approach aligns with broader industry trends, as startups like Monogram and Wabi experiment with generative interfaces that build screens on the fly rather than relying on fixed, pre-designed layouts.

Chesky predicts a hybrid future where deterministic interfaces (predictable and structured) coexist with generative, adaptive designs. This combination aims to provide the reliability users expect from established platforms while offering the flexibility of AI-driven customization.

Building an Agent-Friendly Ecosystem

As AI agents become more prevalent, Airbnb is preparing to become a robust data layer for these tools. Chesky acknowledged concerns that AI intermediaries might cut companies out of direct customer relationships. However, he views AI agents primarily as effective lead generators that can drive users toward booking opportunities.

The company is investing heavily in user interface controls to ensure that agents can operate within Airbnb’s ecosystem without replacing the native app experience. Chesky emphasized that if a user wants features like host messaging, identity verification, or map integration, the agent must either hand off control back to the app or utilize a robust software developer kit (SDK).

"We want to make our app more agentic, and we want to build a lot more user interface controls... either the agent has to do a handoff or the agent application would need to build a bit more of a robust software developer kit."

This strategy ensures that Airbnb retains its brand identity and functional integrity while participating in the emerging agent economy. Chesky rejected the notion of a future where all apps merge into a single universal UI, arguing instead that interfaces should adapt to their specific utility and function.

The Case for an AI Operating System

Perhaps Chesky’s most ambitious claim is that the tech industry lacks a true operating system built for AI. He argues that current AI apps are being layered onto existing platforms like iOS, macOS, and Windows, which were not designed for agent-centric interactions. In his view, AI needs to operate at the kernel level, allowing all components to interact seamlessly.

Currently, the race is underway to become the primary AI agent, or the "quarterback," through which users conduct all digital activities. Chesky suggests this model is flawed because no single entity is developing a comprehensive SDK for this new paradigm. He cited ChatGPT’s early attempt at an app store as an example of misaligned ambitions.

"If you want to be an app store like the iPhone, you need to have the same openness and capability that made the iPhone successful."

Instead of competing to be the sole gateway, Chesky envisions Airbnb becoming an interoperable agent itself. He described a future where the Airbnb explore tab, customer service, and other services are powered by specialized agents. These agents would eventually connect through standards like MCP (Model Context Protocol), enabling seamless communication between different apps and services.

Interoperability Through Agents

Chesky contrasted the lack of interoperability between traditional apps with the potential for agents to bridge gaps. Because agents act as "apps with faces," they can negotiate and interact across platforms in ways static applications cannot. This could lead to a more open internet where different services communicate directly through their respective AI agents.

This thesis implies a fundamental shift in how software interacts. Rather than users manually navigating between disparate apps, their agents could coordinate tasks such as booking travel, ordering food, and managing payments across platforms. For Airbnb, this means positioning its infrastructure as a key node in this emerging network of agent-to-agent interactions.

Conclusion

Airbnb’s latest AI initiatives reflect a company carefully navigating the transition from traditional interfaces to agent-driven experiences. By prioritizing multiplayer functionality, preserving the joy of browsing, and advocating for a dedicated AI operating system, Chesky is laying the groundwork for a more integrated digital future. As the industry grapples with the limitations of chatbots, Airbnb’s deliberate approach may offer a blueprint for balancing technological advancement with user-centric design.