OpenAI is building a humanoid robot, and CEO Sam Altman says the machine’s body is the easier half of the job. Speaking on the Sources podcast in early September, Altman answered a question about whether OpenAI’s hardware push meant a humanoid or a data-centre machine with “Both”, as Forbes reported. “We will definitely do a humanoid,” he said. “We will do other form factors as well.” Engadget’s latest write-up pulls the remarks together with a rough timeline.
That timeline is the newest detail. According to dev.ua’s relay of the interview, Altman told Fortune that OpenAI should have a demo in 2027. Engadget quotes him as saying robots won’t be walking the streets for a few years after that. OpenAI hasn’t said what the demo will show or what the robot will look like, so for now this is a CEO’s target rather than a product plan.
Why a humanoid, and why the brain comes first
Altman’s case for the human shape is practical. On the podcast, he argued that “the world is very much designed for people,” and that doors, keyboards and tools assume a human body. A robot that works in those spaces without modification should probably match that form. Where the body ends up looking is, in his words, “less important than really figuring out … the brain that makes the robot work.”
That’s the harder problem. A useful humanoid has to read its cameras and sensors, interpret instructions, plan the task, coordinate its limbs, and recover when something goes wrong. Picking up a mug means finding the handle, judging the grip, and applying the right force, or you spill. OpenAI has been here before: its Dactyl project trained a robotic hand to handle real objects entirely in simulation, and the company later applied the approach to a Rubik’s Cube. Models can learn the right move in software, but physics still decides whether it works.
The body isn’t an afterthought
Altman’s emphasis on software doesn’t mean OpenAI is ignoring hardware. Engadget’s reading of the company’s robotics listings shows work on custom electromechanical actuators, including motors, transmissions, sensing and thermal systems, plus roles in manufacturing and the supply chain needed to produce hardware at scale. Other roles cover motion control, state estimation, actuator control and the machine-learning infrastructure for training physical AI. Engadget counts 27 robotics openings on OpenAI’s careers page.
The company isn’t alone. Tesla has worked on Optimus for years, and XPENG recently had its IRON humanoid walk off the production line on its own after assembly. Both show how much effort it takes to get a complex set of motors and joints to behave, even with capable software.
The first target isn’t the living room. According to Engadget, one of OpenAI’s robotics postings names general-purpose robots that help skilled workers build infrastructure as the short-term goal, with a personal robot that handles a wider range of tasks as the long-term one. Because a home robot would carry cameras and microphones into private spaces, it will raise privacy questions, though OpenAI hasn’t announced a price, subscription model, release date, or robot-specific privacy framework.
When will OpenAI’s humanoid robot launch?
There’s no release date. Altman told Fortune he expects a demo in 2027 and doesn’t expect robots on the streets for a few years after that, so a consumer product is further out still.
Will the OpenAI robot be for homes?
Not first. OpenAI’s job postings describe infrastructure-building robots that assist skilled workers as the near-term goal, and a personal robot as the longer-term one.
What is OpenAI’s robot “brain”?
It’s the AI that perceives its surroundings, follows instructions, plans tasks, and controls the body. Altman calls it the main challenge, and OpenAI’s robotics roles include motion control, state estimation and training infrastructure for physical AI.
