Chesky is establishing an AI lab, stepping into competition with Altman's OpenAI.

      TL;DR

      Brian Chesky, CEO of Airbnb, intends to support a new AI lab concentrating on user interaction and design, while continuing his role at Airbnb. This places him in competition with Sam Altman, who he helped reinstate at OpenAI in 2023.

      Chesky has been influential in the AI space for years. He met Sam Altman via Y Combinator in 2006, guided him through OpenAI's rapid expansion, and assisted in his reinstatement after the board dismissed him in November 2023. There were discussions about him potentially joining OpenAI’s board.

      Now, Chesky is set to rival his protégé's firm. According to Bloomberg, he plans to support his own AI lab that focuses on user interaction and design, while retaining his position as Airbnb's CEO and not personally managing the lab. The specifics are still in the early stages and may evolve.

      Chesky's dissatisfaction

      This initiative reflects a frustration Chesky has expressed publicly for over a year. Last year, he mentioned that Airbnb had not formed a partnership for large language models (LLMs) because existing products were not ready for his vision. His stance is that travel and commerce necessitate a rich visual interface, rather than the text-based chatbots popularized by OpenAI and Anthropic.

      Airbnb has been active in the AI realm. The company hired Ahmad Al-Dahle as its chief technology officer in January 2026, who had led generative AI projects at Meta, including the Llama model family. They have revamped their app with a large language model for conversational search, automated 40% of customer service queries using an AI bot, and introduced AI-generated listing information and review summaries. A voice-based assistant is also in the works for later this year.

      However, Chesky seems to have determined that relying on AI from leading labs is insufficient. He aims to build at the model layer instead of just at the application layer.

      An emerging trend

      Chesky is not alone in this pursuit. Brett Adcock founded Hark late last year, investing $100 million of his own funds to develop a universal AI interface, then secured a $700 million Series A round at a $6 billion valuation. Hark also focuses on user interaction and hardware, with the lead iPhone designer from Apple now leading its design.

      Mira Murati's Thinking Machines Lab is exploring “interaction models” that process continuous streams of audio, text, and video in real-time. The unifying element is the belief that leading labs have prioritized intelligence over interface, and that the next significant layer lies between the model and the user.

      This trend has broader implications. When founders like Chesky stop waiting for OpenAI, Anthropic, or Google to provide what they seek and start to establish their own research facilities, it indicates that the application layer has reached the limits of what commodity models can offer.

      The Altman factor

      The personal aspect is noteworthy. Chesky and Altman's connection spans nearly twenty years, having met through Y Combinator, which also incubated Airbnb. As OpenAI grew, Chesky began meeting with Altman regularly to provide guidance on scaling a tech company. During the board crisis in November 2023, Chesky advised Altman on public relations and garnered support from Silicon Valley leaders.

      Chesky is now building a venture that will compete, at least in part, with OpenAI's aspirations in user-centric AI. It remains uncertain whether the new lab will develop its own models or create specialized systems using existing ones. However, the direction is evident: Chesky desires proprietary AI research instead of merely an API subscription.

      Questions that remain

      Much about the lab is still unclear. There is no name, team, funding information, or timeline available. Chesky's pledge to stay at Airbnb raises doubts about how much focus the new initiative will receive, and whoever runs it will take over from a founding chair described by TechCrunch as “known as a micromanager.”

      What is clear is the concept. Chesky has closely observed the AI lab landscape and concludes that the interface challenge—making AI effective in rich, visual, consumer-oriented settings—is significant enough to justify a dedicated research operation. Whether a part-time founder can create something impactful remains the unanswered question.