The US Is Blocking New Foreign-Made Humanoid Robots Without Security Approval

The United States has placed a major new barrier in front of foreign-made humanoid and four-legged robots. From July 28, new models produced outside the country generally cannot receive the federal equipment authorization needed to be imported, marketed or sold, unless they pass a separate security review.

KEY FACT
The restriction applies to new foreign-produced models seeking authorization, not to robots already bought or models that were previously approved.

What the FCC actually changed

The Federal Communications Commission (FCC) maintains a Covered List of communications equipment judged to pose an unacceptable risk to US national security. Its July 28 update added advanced robotic devices made in foreign countries, along with foreign-produced connected power inverters.

The robot category includes mobile machines such as humanoids and quadrupeds. Because connected robots contain radio equipment, a new model normally needs FCC equipment authorization before it can be imported or marketed in the United States. Placement on the Covered List generally closes that route.

The wording is broader than a ban aimed at named Chinese companies. It applies according to where a device is produced. Chinese manufacturers are likely to feel the greatest immediate effect because they have become prominent suppliers of affordable humanoid and four-legged platforms, but models made in other foreign countries are also within the rule.

Why a robot is being treated as a security risk

An advanced robot is not merely a motorized appliance. It may carry cameras, microphones, depth sensors and mapping systems, then send data through wireless networks. In a factory, laboratory or power facility, that sensor package could reveal the layout of a site, the movement of workers or details about equipment and operations.

The FCC also points to the possibility of remote access or commandeering. A compromised robot can create a physical risk as well as a data risk because it can move, lift objects and operate near people. The concern is therefore different from an ordinary connected gadget, even though the public notice does not show that every foreign robot is unsafe.

Reuters highlighted Unitree, a Chinese company known for humanoid and quadruped robots, as one manufacturer likely to be affected. Its G1 humanoid is marketed for research and development. The FCC rule, however, does not accuse that specific model of a demonstrated vulnerability.

What remains allowed

The change does not order owners to stop using robots they already have. Previously purchased devices are unaffected, and equipment that already received authorization does not automatically lose it. The federal government can also continue buying or using covered devices.

There is a route for new foreign-made models. A robotics manufacturer can seek Conditional Approval from the US Department of War. If that department concludes that a device does not pose an unacceptable national-security risk, the FCC can allow it to proceed through equipment authorization.

This exception matters because it changes the policy from an absolute prohibition into a security gate. How quickly reviews happen, what technical evidence companies must provide, and whether approvals apply to individual models or product families will determine how restrictive the system becomes in practice.

Why this could change the robotics market

Humanoid robots are still early products, but companies are testing them for logistics, manufacturing, inspection and research. Affordable foreign platforms have given universities and developers a faster way to experiment. If newer models cannot reach the US without a lengthy review, buyers may face fewer choices and higher costs.

The policy could also encourage assembly and production inside the United States, although domestic manufacturing alone will not settle every cybersecurity question. Software updates, cloud services, component origins and data-handling practices can matter as much as the final assembly location.

For manufacturers, the next contest will be about evidence as well as price and performance. Companies may need to document secure update systems, data flows, remote-access controls and supply chains before a robot can enter one of the world’s largest technology markets.

What happens next

Robot makers will have to decide whether to pursue federal approval, shift production, redesign connected features or focus on markets outside the United States. US importers and research institutions will also need to check the authorization status of future models before ordering them.

The larger takeaway is that robotics policy has moved beyond questions about jobs and safety on the factory floor. Governments are beginning to treat intelligent machines as mobile networks of sensors and actuators, and access to markets may increasingly depend on proving that those machines can be trusted.