📅 ⏱ 4 min. read

Meta Acquires Humanoid Robotics Startup to Challenge Tesla, Nvidia, and Amazon in the Physical AI Race

Meta Platforms has acquired Assured Robot Intelligence in its first major robotics acquisition. The deal brings renowned robotics researchers Xiaolong Wang and Lerrel Pinto into Meta’s Superintelligence Labs as the company positions itself against rivals including Tesla, Amazon, NVIDIA, and Alphabet in the race to build the dominant physical AI platform.

Meta Acquires Humanoid Robotics Startup to Challenge Tesla, Nvidia, and Amazon in the Physical AI Race

Meta Platforms has acquired San Diego-based humanoid robotics startup Assured Robot Intelligence (ARI) in a strategic move that significantly expands the company’s ambitions beyond social media, virtual reality, and digital AI systems. The acquisition marks Meta’s first major external robotics purchase and positions the company directly inside the rapidly accelerating race toward “physical AI” and embodied intelligence.

While financial details were not disclosed, the entire ARI team — including co-founders Xiaolong Wang and Lerrel Pinto — will join Meta’s Superintelligence Labs (MSL), strengthening Meta’s capabilities in humanoid robotics, robot learning, and adaptive AI systems.

The deal comes shortly after Meta’s reported $2 billion attempt to acquire Manus AI was blocked by Chinese regulators, pushing the company toward alternative paths to establish a foothold in the emerging humanoid robotics market.

Meta’s Push Into Physical AI

ARI specializes in foundation models for humanoid robots — large-scale AI systems designed to help machines understand, predict, and respond to human behavior inside real-world environments. The company’s research focuses heavily on domestic and unstructured settings where robots must continuously adapt to changing surroundings.

This acquisition gives Meta immediate access to advanced research in:

  • Whole-body humanoid robot control
  • Self-learning robotics systems
  • Adaptive manipulation
  • Embodied AI reasoning
  • Human behavior prediction models
  • Physical environment interaction

Industry analysts increasingly believe that embodied intelligence — AI systems capable of physically interacting with the real world — is a necessary step toward artificial general intelligence (AGI).

Meta appears to be positioning itself not as a humanoid hardware manufacturer alone, but as the builder of the “core intelligence layer” powering future robotic ecosystems. Several analysts have compared this strategy to Android’s role in smartphones: a foundational software platform used across many hardware vendors.

The Robotics Talent Behind ARI

The acquisition also delivers a highly influential robotics research team into Meta’s AI division.

Xiaolong Wang

ARI co-founder Xiaolong Wang previously worked as a robotics researcher at Nvidia and is widely recognized for his work in robot learning, computer vision, and autonomous behavior systems.

Lerrel Pinto

Co-founder Lerrel Pinto previously co-founded Fauna Robotics, a social robotics startup that was acquired by Amazon earlier in 2026. Pinto’s work focuses on scalable robot learning and autonomous adaptation in real-world environments.

The combination of Wang and Pinto gives Meta one of the strongest robotics AI research teams currently assembled inside a major technology company.

Meta’s Main Competitors in the Humanoid Robotics Race

Meta now enters an increasingly crowded and aggressively funded humanoid robotics market dominated by both Big Tech firms and specialized robotics startups.

Tesla

Tesla remains one of the most closely watched players in humanoid robotics through its Optimus program. Earlier this year, the company reportedly shifted portions of its manufacturing infrastructure away from certain vehicle programs to accelerate Optimus production capacity.

Tesla’s strategy focuses heavily on vertically integrated humanoid hardware and manufacturing automation.

Nvidia

Nvidia is emerging as a dominant “AI brain” provider for humanoid robots through its GR00T robotics foundation model initiative.

Rather than manufacturing robots directly, Nvidia is building the underlying AI stack and simulation infrastructure used by robotics manufacturers worldwide.

Amazon

Amazon has aggressively expanded its robotics ambitions through acquisitions and talent recruitment.

In March 2026, Amazon acquired Fauna Robotics — co-founded by ARI’s Lerrel Pinto — and also hired the founding team behind Covariant, a robotics AI startup focused on warehouse automation and robotic reasoning systems.

Alphabet (Google)

Alphabet continues expanding its robotics ecosystem through Intrinsic and Gemini Robotics.

Google has also partnered with Boston Dynamics and invested in humanoid robotics startup Apptronik as it pushes deeper into industrial robotics and factory automation.

Microsoft

Microsoft remains heavily involved in robotics infrastructure and AI foundation models through projects such as Rho-alpha while also providing Azure AI infrastructure to several robotics companies.

Specialized Humanoid Robotics Companies

Alongside major technology firms, Meta must also compete against specialized humanoid robotics companies rapidly moving toward commercialization.

Figure AI

Figure AI has become one of the highest-profile humanoid startups globally, backed by Nvidia, OpenAI, Microsoft, and Jeff Bezos. The company is reportedly targeting 100,000 humanoid robot deployments within four years.

Boston Dynamics

Owned by Hyundai Motor Group, Boston Dynamics recently introduced a fully electric successor to its Atlas humanoid robot platform, signaling a shift toward scalable industrial deployment.

Xiaomi

Xiaomi has already begun deploying humanoid robots inside electric vehicle assembly plants where they reportedly achieved high autonomous task success rates.

Sanctuary AI

Sanctuary AI continues expanding industrial deployment partnerships, including collaborations with Magna International for automotive manufacturing applications.

The Emerging Physical AI Arms Race

The acquisition of ARI signals that the competition for humanoid robotics leadership is rapidly shifting from experimental research into a large-scale strategic technology battle.

Unlike earlier robotics waves focused primarily on hardware engineering, the next generation of humanoid systems is increasingly centered around foundation AI models, self-learning behavior systems, and embodied reasoning capabilities.

Meta’s strategy suggests the company believes the long-term value lies not only in building robots, but in owning the intelligence layer that powers them.

If successful, Meta could position itself at the center of a future ecosystem where humanoid robots operate across homes, factories, logistics centers, and consumer environments using AI models developed and controlled by Meta itself.

The acquisition of ARI may ultimately represent the beginning of Meta’s transition from a digital platform company into a major force in physical AI infrastructure.