OpenAI shifts to general AI models directly controlling robots, no longer developing specialized models separately.

November 14, according to Fortune magazine, OpenAI has disbanded its independent robotics research team and reassigned the relevant researchers to other departments within the company. OpenAI stated that this move is not an abandonment of its robotics business, but rather to accelerate the development of general-purpose robots. After the adjustment, robotics researchers will be merged into the core research team led by Chief Research Officer Mark Chen, while the applied robotics and hardware teams will move to the applied division headed by President Greg Brockman.

This adjustment reflects a shift in OpenAI's strategic direction: the company believes that general-purpose AI models can already be directly applied to robot control, eliminating the need to specifically develop independent robotics AI models. OpenAI stated that frontier general-purpose models, including GPT-6Astra, have already demonstrated robotics application capabilities. A study by Nvidia shows that agents equipped with Astra achieved a success rate of 74.8% in simulated robotic operation tests, higher than the 51% previously reported by specialized robotics models on the same leaderboard.

OpenAI still plans to develop humanoid robots, and CEO Altman previously proposed aiming for a first public demonstration in 2027. However, industry insiders point out that general-purpose models still have limitations such as slower running speeds and reliance on cloud networks, and robotics companies still hold advantages in real-world deployment and data accumulation.