The FCC just redrew the map for humanoid robotics, and almost nobody outside the industry has clocked how big this is.
This week the FCC added advanced robotic devices such as humanoid and quadruped robots plus connected power inverters to its Covered List, the roster of foreign-made gear deemed too risky to authorize for U.S. sale. Chairman Brendan Carr framed it as supply-chain security, “acting in lock step with our national security agencies.” The mechanism is simple and total: without FCC equipment authorization, a device cannot be imported, marketed, or sold in the United States. No authorization, no market.
Here’s why that matters more than the headline suggests.
China doesn’t have a foothold in humanoid robotics. It has the market. Six of the top ten manufacturers by shipment volume are Chinese, and together they account for 87% of all humanoid robots shipped globally with Unitree, Agibot, and UBTech leading the pack. Unitree just landed commercially in Europe on July 22, its first Western market entry, with a U.S. push reportedly next. This order lands directly in front of that. So much for a Unitree IPO…that’s pretty much DOA.
Three things I’m watching, because they’ll determine whether this is a real shift or a paper tiger:
- Conditional Approval mechanics. There’s a conditional approval path, the Department of War for robots, and DHS for Inverters, for companies that can show they don’t pose unacceptable risks due to their secure supply chains and locked-down firmware. Nobody knows the review timeline or evidentiary bar yet. If it’s fast and workable, this is a speed bump. If it’s slow or effectively unusable, it’s a wall.
- The order is forward-looking, not retroactive. Already-authorized models can keep selling. This is a gate on new launches, not a recall. That distinction matters for anyone modeling near-term revenue impact for Chinese OEMs already in-market.
- Whether U.S. players can actually fill the vacuum in volume. American humanoid companies such as Figure, 1X, Agility Robotics, and Apptronik among them, now get a protected runway, backed by an estimated $150B in aggregate capital raised across roughly 180 U.S. companies working on humanoids and components. Protection is not the same as readiness. Shipment volumes industry-wide are still small. The order removes a competitor; it doesn’t manufacture capacity.
I believe this is a genuine structural move, not a symbolic one, but its ultimate size depends entirely on execution details we don’t have yet. I’d treat “how strict is Conditional Approval in practice” as the single most important variable to track over the next two quarters.
The deeper issue: a Covered List entry doesn’t build a factory.
Here’s the part I think gets skipped in the celebration of this order. The FCC just closed a door. It did not open a supply chain. And the gap between those two things is the real story.
China’s advantage in humanoid robotics was never just labor cost, it’s component control. Industry mapping puts China’s share of key humanoid-robot component manufacturing at roughly 63%, with dominance concentrated in actuators, reducers, servo motors, and batteries. The deeper chokepoint is materials: China controls somewhere around 90% of heavy rare earth processing globally and closer to 99% of processing for dysprosium and terbium specifically, the elements that go into the high-strength permanent magnets every electric motor in a humanoid robot depends on. You cannot manufacture a competitive humanoid at volume in the U.S. today without touching that processing chain somewhere.
So “move manufacturing back to the USA” is really three separate reshoring problems stacked on top of each other, on three different timelines:
1. Final assembly — already happening, and further along than most people think. Figure is building actuators, hands, and batteries in-house rather than sourcing from China, and is targeting production capacity in the hundreds of thousands of units annually. 1X is expanding toward 100,000+ annual capacity with a second California facility in the pipeline. Apptronik is building in Texas. Agility Robotics runs its RoboFab facility in Salem, Oregon. Standard Bots, which just raised $200 million, says it plans to manufacture everything domestically. This is the layer where reshoring is real, funded, and underway… not aspirational.
2. Precision components — much harder, and where the real bottleneck sits. This is not a “build a new factory” problem, it’s a “rebuild two decades of process knowledge” problem. Historically, two Swiss companies held more than half the global market for planetary roller screws. China closed that gap through reverse engineering and years of capacity investment, not overnight. The U.S. doesn’t currently have that manufacturing base at scale, and industry analysis is blunt about why: the equipment to build these components is globally available, but the accumulated process expertise, what one industry report called “manufacturing software,” is not something capital alone reconstructs quickly.
3. Rare earth materials — the slowest-moving layer, and arguably the most strategically important. This is where the federal government has actually put real money down, not just rhetoric. The Department of War has entered a multibillion-dollar public-private partnership with MP Materials to build a second domestic magnet facility (the “10X” plant), on top of Mountain Pass in California and the existing Independence facility in Fort Worth, TX, which began neodymium-iron-boron permanent magnet production in December 2025 and is expected to reach commercial magnet sales in the second half of 2026. Combined U.S. capacity across current and planned facilities could reach roughly 30,000 metric tons annually by 2030. That sounds significant until you register that heavy rare earth processing of dysprosium and terbium, remains largely untouched by these projects. Light rare earth reshoring is happening. Heavy rare earth reshoring barely exists yet, in the U.S. or anywhere in the West.
Why the labor argument is weaker than people assume. The old objection to reshoring of “American labor is too expensive,” is increasingly beside the point, because the plants being built don’t compete on labor cost anymore. They compete on automation. Reshoring Initiative data shows roughly 244,000 announced manufacturing jobs returned to the U.S. in 2024, with 88% of those classified as high-tech or medium-high-tech engineering and skilled-technician roles, not assembly-line headcount. A highly automated U.S. plant running collaborative robots and machine-vision inspection can now match unit economics with an overseas facility on labor alone. The binding constraint isn’t wage rates. It’s a reported shortfall of roughly 400,000-plus open U.S. manufacturing jobs against a shrinking qualified labor pool, which is its own irony, since the industry these plants serve is the one supposed to solve exactly that problem.
Conclusion
I strongly advise investors and robotics companies located in the U.S. to take full advantage of the ban. This means doing more than making silly videos of your dainty robots unloading dishwashers. This is an opportunity to manufacture the infrastructure required to dominate this industry. This is the time to build humanoid robots that can legitimately do the “hard and dirty jobs” that many humanoid robotics company CEOs claimed they were building. By :hard and dirty,” I mean build a humanoid robot that can change and mount car and truck tires using the same tools and process as the associates at Discount Tire, and mechanics in shops all over the USA currently utilize for changing and mounting tires.
This is especially the time for America’s leading robotics companies to build humanoid robots that take lethality on the battlefield to the next level.
I don’t claim that this will be easy. What I can state is that there are companies like Chang Robotics, led by Matthew Chang and Kate McAfoose, with the skills and experience to design factories for automation, select and install the best automation, accelerate the manufacturing of robots and components, and design and implement supply chain and logistics strategies that will enable growth. One more interesting fact…Chang Robotics is also a leader in rare earths minerals.
Now isn’t the time for celebrating, it’s the time for building. Get started.