Meta has completed the acquisition of Assured Robot Intelligence (ARI), a startup focused on humanoid robotics, for an undisclosed amount, the company confirmed. The ARI team, including its co-founders, will be integrated into Meta’s Superintelligence Labs research division within its AI unit. The startup had previously raised a seed round of undisclosed size from AIX Ventures, a firm specializing in early-stage AI investments.

ARI was developing foundation models designed to enable humanoid robots to perform a wide range of physical tasks, including household chores. Co-founder Xiaolong Wang previously worked as a researcher at Nvidia and served as an associate professor at UC San Diego, earning numerous prestigious awards throughout his career. Co-founder Lerrel Pinto, who previously taught at NYU and co-founded the kid-size humanoid startup Fauna Robotics—acquired by Amazon last month—also boasts an impressive list of accolades. Meta plans to leverage ARI’s expertise to advance its humanoid robotics ambitions. “This team, led by Lerrel Pinto and Xiaolong Wang, will bring a deep expertise in how we can design our models and frontier capabilities for robot control and self-learning to whole-body humanoid control,” the company stated.

Meta researchers have been exploring humanoid robotics technology for several years. A leaked internal memo from about a year ago outlined Meta’s ambitions to develop such a robot, including both AI models and hardware, with a consumer focus in mind. Even if Meta never releases a commercial humanoid product, many AI experts now believe that achieving artificial general intelligence (AGI)—the theoretical point where AI matches or surpasses human intelligence across all domains—will require training AI models in the physical world, where robots learn through direct interaction rather than relying solely on data.

The acquisitions of ARI and Fauna reflect a broader industry push toward humanoid robotics, with market forecasts varying widely. Goldman Sachs projects the market could reach $38 billion by 2035, while Morgan Stanley estimates it could hit $5 trillion by 2050. This wide spread highlights both the enormous potential and the significant uncertainty surrounding a technology that is still finding its footing.