Meta is moving deeper into humanoid robots, with Bloomberg reporting that the company has begun work on both hardware and software for AI-powered machines. The effort sits inside Reality Labs and starts with a practical target: robots that can handle everyday household tasks.
A platform play, not just a robot project
According to Bloomberg, Meta has assembled a dedicated Reality Labs team for the work. The goal is not simply to build a single finished robot and keep the technology inside Meta. Instead, the company aims to create fundamental building blocks that other companies could use to make and sell their own robots.
Those building blocks include AI systems, sensors, and software. That approach matters because humanoid robots require more than a physical frame. They need perception, control, decision-making, and the ability to understand spaces that were designed for people rather than machines.
The household focus also points to the difficulty of the challenge. Everyday tasks can look simple to a person but require a robot to recognize objects, move safely, and respond to changing surroundings. Meta’s reported strategy suggests it wants to provide the underlying technology for that problem, even if other manufacturers handle the final products.
Robotics partners are already part of the picture
Bloomberg reports that Meta has started discussions with robotics manufacturers. Those talks include Unitree Robotics and Figure AI.
Figure AI is a notable name in the source report because it recently ended its partnership with OpenAI, stating that LLMs are now a commodity. That detail shows how quickly the robotics field is shifting around AI capabilities. Companies working on humanoid robots are not only choosing hardware partners; they are also deciding which software and model layers are distinctive enough to build around.
For Meta, the opportunity described in the report is to become useful to companies that already know how to manufacture and sell robots. If Meta can supply AI systems, sensors, and software, it could participate in the humanoid robotics market without needing to be the only company producing the finished machines.
Reality Labs may give Meta a useful base
Bloomberg reports that Meta believes its experience in virtual and augmented reality gives it an advantage. That history matters because AR and VR work also depends on understanding space, movement, sensors, and how digital systems interpret the physical world.
The source article notes that data collected from Meta’s AR and VR devices could accelerate robot development significantly. Meta has been collecting data using its own devices specifically for AI training since at least 2022, when it released the Project Aria pilot dataset.
That connection helps explain why Reality Labs is central to the project. The same division associated with virtual and augmented reality may have technical assets that are relevant to robots operating in homes. A humanoid robot needs to interpret rooms, objects, and human-scale environments. AR and VR systems also depend on mapping and understanding those kinds of environments, even though the end product is different.
The humanoid robot race is getting crowded
Meta is not entering an empty field. The source article says Google DeepMind and OpenAI are actively working on similar robotic hardware and software projects. It also points to Elon Musk with Optimus, powered by Tesla and xAI, as well as several startups.
This broader activity reflects a larger shift in AI robotics. Recent advances in large multimodal models have helped make standard robots more versatile by improving their understanding of the world around them. The source article cites Figure AI as an example of that progress.
Large multimodal models matter in this context because robots need to process more than one kind of information. A machine working in a home has to connect visual understanding, language, physical movement, and changing context. The source does not claim that this problem is solved, but it does make clear that advances in these models are helping robots become more adaptable.
The hard questions are still unresolved
Even with Meta’s reported investment and the wider push from major AI companies, the source article emphasizes that important questions remain. The biggest issue is how humanoid robots will fit into everyday life.
That question is not only technical. A robot built for household tasks must be useful enough to justify its place in the home. It must also work around people, furniture, routines, and unpredictable situations. The source article also notes that hardware costs are currently high, which remains a major barrier for the category.
Meta’s reported plan is therefore best understood as an early move into an emerging market rather than a finished consumer moment. The company is building AI-powered humanoid robot technology, exploring partnerships, and trying to use its Reality Labs experience as leverage. Whether that becomes a widely used household technology will depend on how well these systems handle real environments and whether the cost problem can be addressed.