Chesky is planning an AI lab to compete with Altman’s OpenAI

Brian Chesky has spent years as the king of artificial intelligence. He met Sam Altman through Y Combinator in 2006, advised him on managing OpenAI’s hyper-growth, and helped broker Altman’s return to power after the board fired him in November 2023. He is slated for a seat on OpenAI’s board.

Now he is competing with his patron’s company. Chesky plans to support his own new AI lab, Bloomberg reported first Wednesday, with a focus on user interaction and design. He will remain CEO of Airbnb and will not lead the lab himself. Details are preliminary and subject to change.

Why is Chesky unhappy?

The move reflects a frustration that Chesky has voiced publicly for more than a year. He said last year Airbnb did not enter the LLM partnership the existing products were not quite ready for what he wants to build. His argument is that travel and commerce require a rich visual interface, not the text-based chatbots popularized by OpenAI and Anthropic.

Airbnb has not been idle in artificial intelligence. The company hired Ahmed Al-DahleAs CTO in January 2026, he led the generative artificial intelligence work at Meta, including the Llama model family. It rebuilt its program around a large language model for conversational search, automated 40% of customer support inquiries with an AI bot, and introduced AI-generated listing details and overview summaries. It is planned to create a voice assistant at the end of this year.

But Chesky concluded that acquiring artificial intelligence from frontier labs was not enough. It wants to build at the model level, not just the application level.

A growing model

Chesky is not alone. Brett Adcock launched Hark raised $100 million of its own money, then a $700 million Series A at $6 billion, late last year to build a universal AI interface. Hark also emphasizes user interaction and hardware, with Apple’s lead iPhone designer now leading the design effort.

Mira Muratini Laboratory of thinking machines operates on “interaction models” that process continuous streams of audio, text, and video in real-time. A common thread is the belief that frontier labs focus on intelligence at the expense of the interface, and that the next defensible layer sits between the model and the user.

The trend has a wider meaning. When Chesky’s founders stopped waiting for OpenAI, anthropicor deliver what they need to Google and start building their own research capacity, indicating that the application layer has reached the limits of what commodity models can provide.

Altman dynamics

The personal dimension is hard to ignore. Chesky and Altman have been in a relationship for nearly two decades. Chesky met Altman through Y Combinator, which incubated Airbnb. When OpenAI launched, Chesky began meeting regularly with Altman to advise on scaling the tech company. During the November 2023 board crisis, Chesky provided Altman with public relations advice and garnered support among Silicon Valley executives.

Now Chesky is building an operation that will at least partially rival OpenAI’s own ambitions in user-facing AI. It is unclear whether the new lab will train its own models or build custom systems on top of existing ones. But the direction is clear: Chesky wants proprietary AI research, not API subscriptions.

Things we don’t know

Almost everything about the lab remains unspecified. There is no name, no announced team, no announced funding and no timeline. Chesky’s commitment to stay at Airbnb raises questions about how long the new venture will hold his attention and whether the person leading it will inherit the founding chair, which TechCrunch described as “”known as a micro manager.“

What is clear is the thesis. Chesky followed AI lab landscape closer than almost anyone outside, and he decided that the interface problem that makes AI useful in rich, visual, consumer-related contexts was important enough to warrant his own research operation. Whether a part-time founder can build something important is an open question.