OpenAI is building a humanoid robot, chief executive Sam Altman has confirmed. Asked on a podcast whether the company would make one, he said it would. A human form makes sense, he argued, because the world is designed for people, and matching that form factor seems good. The harder problem, he said, is the brain that makes the robot work, not the body.
In a separate interview he put a date on it: a demo in 2027. He added that he does not expect robots walking the streets for a few years longer.
The company's recruiting backs up the statement. Its robotics roles cover motion control, state estimation, actuator control, whole-robot behaviour and machine-learning infrastructure for physical AI. Other postings seek engineers for custom electromechanical actuators, including motors, transmissions, sensing and thermal systems, and for manufacturing and supply chains for future hardware at scale. One posting describes a short-term goal of general-purpose robots that help skilled workers build infrastructure, and a long-term goal of a personal robot for a wider range of tasks. The careers page lists 27 robotics openings.
OpenAI has done robotics research before. An earlier project trained a robotic hand to manipulate real objects using simulation, a technique later applied to solving a Rubik's Cube. It now enters a field where Tesla has been developing its Optimus humanoid for years and where XPENG's IRON humanoid recently walked off a production line.
Much remains undisclosed: what the 2027 demo will show, what the robot will look like, when a product might launch, what it will cost, when outside users might get access and how privacy will be handled. No robotics leaders or hardware and manufacturing partners have been named. The company is also working on new consumer hardware, with no details given.
