OpenAI has disbanded its stand-alone robotics team, moving its researchers into other parts of the company in a bet that its general-purpose AI models can now run robots directly — no dedicated team building robot-specific “brains” required.
The company confirmed the reorganization to Fortune, which reported the move on November 13. A spokesperson said robotics remains a core part of OpenAI’s long-term vision and framed the change as a commitment to the work rather than a retreat from it.
The robotics research team now sits inside OpenAI’s main research organization, reporting to Chief Research Officer Mark Chen, while the applied robotics and hardware teams have been folded into the larger Applied organization, which reports to President Greg Brockman.
The benchmark that changed the plan
The trigger, according to Fortune, was a striking result: an agent powered by OpenAI’s GPT-6 Astra model scored 74.8% on a simulated robot manipulation benchmark run by Nvidia researchers, compared with 51.0% for the best specialized robotic models on the same leaderboard.
Researchers were surprised by how well a general-purpose frontier model performed when connected to robots, Fortune reported. At the start of 2026, OpenAI’s researchers had assumed the best route to capable robots ran through models like Sora, its video generator, which has to grasp something of how the physical world works to make realistic clips. GPT-6 Astra went further, applying directly to physical tasks.
Outside experts were more cautious. Guanya Shi, an assistant professor at Carnegie Mellon University’s Robotics Institute, and Russ Tedrake, an MIT professor and chief executive of Walden Robotics, both noted that simulated benchmarks do not settle the question of real-world performance, where latency, hardware and messy physical environments still matter. Pete Florence, co-founder of Generalist, a startup training general-purpose models for robots, said it was unclear any single model would end up dominating robotics.
A second ending for OpenAI robotics
This is the second time OpenAI has closed a robotics team. The company wound down its original unit in 2020, with co-founder Wojciech Zaremba later citing limited robot data and a more efficient path to artificial general intelligence through other methods.
OpenAI rebuilt a stand-alone team in 2024 and announced a partnership with humanoid startup Figure to develop AI models for humanoid robots. That partnership ended in February 2025, and the team lasted roughly 18 months in its second incarnation.
The move came so recently that OpenAI had already registered a booth at CoRL, a leading academic robotics conference held in Austin this week. Fortune reported the booth stood empty when its reporter visited.
A humanoid is still on the way
OpenAI says it is still developing its own humanoid robot, with chief executive Sam Altman targeting a first public demonstration in 2027. The company has not published a blog post about the change and has not said how many people were affected.
The reorganization lands while OpenAI is under pressure to cut costs and narrow its focus. Fortune reported operating losses of $10.9 billion in the most recent quarter as the company seeks to close a new funding round and prepares for a possible initial public offering in early 2027.
The bet, in short, is that the robot brain is no longer a separate product — it is the same general model, applied to a new body.
