Airbnb has introduced AI-powered search as part of its fall product update, while CEO Brian Chesky says chatbot-only interfaces still do not work well for travel discovery. The company is instead testing AI that works alongside browsing, visual controls, voice, and eventually other agents.
The new AI search features let users search homes, experiences, and services using natural language or voice. Airbnb can also generate relevant filters, highlight listing features, and compare homes from a user’s wishlist with AI-generated summaries.
Chesky told TechCrunch that he still does not consider a chatbot the right interface for shopping or travel because users receive only a few options at a time and often need several exchanges to reach a result. He also does not consider Airbnb’s current AI search interface its final form.
Airbnb Wants AI To Work With Browsing
Chesky said travel planning differs from tasks where users simply want an agent to complete an action. People often enjoy browsing destinations and accommodations, so Airbnb does not want AI to remove the visual discovery process.
He also sees a limitation in AI tools designed around one person. Over the next three to six months, Airbnb plans to study “multiplayer” AI interfaces that could allow several people, such as friends planning a trip together, to interact with the same AI experience.
Chesky expects future software to combine predictable, pre-designed interfaces with generative elements created by AI. He also wants Airbnb to become more compatible with consumer agents such as Muse and Instinct, which could direct users toward Airbnb before handing them into a richer interface for maps, comparisons, host messaging, identity verification, and booking.
Airbnb Prepares For Agent-To-Agent Communication
Airbnb is also considering how its own agents could communicate with agents from other services. Chesky said the Explore tab, customer support, and other areas could each have dedicated agents before eventually becoming part of a larger Airbnb agent.
That system could connect with outside agents through standards such as Model Context Protocol, or MCP. Chesky described agents as a possible way to make services more interoperable without requiring the traditional business integrations that apps often need.
He argues, however, that current consumer AI lacks the software infrastructure required to replace apps completely. Chesky said a proper AI operating system would need richer software development tools, interoperability between agents, and interfaces suited to the functions users are trying to perform.
As long as Apple and Google continue controlling the major mobile platforms, he expects apps to remain important. Without a new platform layer, Chesky expects companies to put agents inside existing apps rather than replace the entire app with an interoperable agent.
Voice Agents Are Coming To Airbnb
Voice will also become part of Airbnb’s AI interface. Chesky expects people to speak to computers more frequently and said Airbnb plans to introduce voice agents for search and customer service this fall.
Airbnb has already expanded AI across customer support, with its Q2 2026 update saying nearly 45% of issues that begin with its AI assistant are resolved without a human agent. Chesky also said he personally uses AI to retrieve information faster, reducing the number of meetings he needs to attend.
Featured image credits: Wikimedia Commons
For more stories like it, click the +Follow button at the top of this page to follow us.