cvtoken.vip

America’s AI Trade War Expands To Humanoid Robots

TOPSHOT - A robot crosses the finish line in the humanoid robot half marathon in Beijing on April 19, 2025. Dozens of humanoid robots took to the streets of Beijing, joining thousands of their flesh-and-blood c...

TOPSHOT-CHINA-ATHLETICS-TECHNOLOGY-ROBOTS

TOPSHOT - A robot crosses the finish line in the humanoid robot half marathon in Beijing on April 19, 2025. Dozens of humanoid robots took to the streets of Beijing, joining thousands of their flesh-and-blood counterparts in a world-first half marathon showcasing China's drive to lead the global race in cutting-edge technology. (Photo by Pedro PARDO / AFP) (Photo by PEDRO PARDO/AFP via Getty Images)

AFP via Getty Images

The United States spent years trying to control the flow of advanced computer chips to China, but now Washington is now turning its attention to AI that’s coming into our borders. On July 28, the Trump administration moved to block new foreign made advanced robots and connected power inverters from entering the American market. The Federal Communications Commission placed the equipment on its Covered List, a regulatory tool reserved for technologies judged to pose an unacceptable risk to national security.

The rules largely target Chinese manufacturers. They cover new models of humanoid robots, four legged robots and other connected machines that combine sensors, software and wireless communications. Previously authorized products can remain on sale, though the FCC retains power to revisit those approvals. New covered devices will struggle to obtain the equipment authorization needed for commercial sale in the United States.

AI Policy Enters The Physical World

Much of the public policy debate about AI leadership has centered on semiconductors, data centers and large language models. While those technologies supply the digital computing brainpower, it’s physical systems such as robots that give the software hands, legs, cameras and access to buildings.

Besides providing autonomous and automated labor to help with various tasks, AI-powered robots can collect spatial data inside a factory. It can map equipment, observe workers, manipulate tools and send information to a remote server. Robots can perform dangerous, dirty or dull tasks that otherwise people would do such as in tunnels, warehouses, military facilities or damaged infrastructure that would be difficult for humans to reach.

That makes connected robots a different security proposition from chatbots. A foreign language model may have privacy, moderation and confidentiality impacts but a networked robot ha more impact on the physical space in which we live and work. It can observe a workplace, touch machinery and move through restricted space.

MORE FOR YOU

FCC officials cited fears involving surveillance, data theft, remote interference and cyberattacks. U.S. policymakers have voiced parallel concerns about connected vehicles, drones, telecommunications equipment and Chinese made port cranes. Robotics now joins that widening category of products whose commercial functions cannot be cleanly separated from their data and communications systems. In addition, connected power inverters, which convert electricity produced by solar panels and batteries into usable current, face comparable restrictions.

AI Policy Shifts

Congress had already been moving in this direction. A bipartisan group of lawmakers introduced the Guarding the U.S. Against Adversarial Robotics Dominance Act in June. The GUARD Act would require federal scrutiny of robots produced by China and other foreign adversaries, and prohibit imports found to threaten national security. A provision requiring the Pentagon to begin such a review later passed the House as part of its defense authorization bill.

The restrictions show how Washington’s definition of AI infrastructure is changing. A few years ago, the phrase referred mainly to data centers, graphics processors and high speed networks. The definition now reaches into power generation, industrial controls, batteries, motors, sensors and machines capable of acting on an AI model’s instructions.

This reflects the broader shift in AI away from just digital systems to physical AI. AI model and infrastructure companies are racing to build models that can reason about images, sound and movement. Robot makers are pairing those systems with machines designed for warehouses, factories, hospitals and homes.

The result is often called embodied AI or physical AI. The term describes systems that learn from the real world and take action inside it. Training a robot to handle thousands of unpredictable tasks requires far more than a clever mechanical frame. It demands data from cameras and force sensors, simulation tools, control software, computing power and repeated operation in real settings. Each deployed machine can become a data collection node. A large robot fleet can produce immense records of how people, tools and industrial environments behave.

That feedback loop may become one of the sector’s strongest competitive moats. The company that deploys the most robots could gather the richest operational data. Better data could produce better models. Better models could support safer and cheaper machines, which could then generate still more data. The cycle resembles the advantage that large internet platforms built from clicks, searches and social activity. This time, the raw material comes from warehouses, sidewalks and assembly lines.

China’s Emerging Lead In Robotics

The International Federation of Robotics estimates that China operates about 2 million industrial robots, roughly 4.5 times the stock in Japan, the next largest market. China accounted for 54% of industrial robot installations worldwide in 2024, and its factories installed 295,000 units that year, around ten times the estimated U.S. total.

Chinese manufacturers are taking a larger share of that home market too. In 2020, domestic suppliers accounted for 30% of installations. By 2024, their share had climbed to 57%, according to the federation. Beijing has made clear that it wants that momentum to continue. Its latest five year plan treats robotics as a core industrial priority and pushes AI research toward machines that can work, move and make decisions in the physical world.

Today humanoid robots remain immature. Many public demonstrations involve controlled settings, trained movements or narrow tasks. Conventional industrial robots still beat humanoids on speed, precision and reliability for repetitive factory work. But in the long term, continued progress and real-world implementations matter more than the current product demonstrations.

Chinese companies can draw on dense networks of battery suppliers, motor manufacturers, electronics plants and precision component makers. They can test designs against one of the world’s largest manufacturing markets. Lower production costs could let them place more machines in real workplaces, where those machines generate the data needed to improve.

Unitree, one of China’s best known robotics companies, sells humanoid and quadruped robots at prices that have attracted laboratories, developers and commercial buyers abroad. In an initial public offering filing, Unitree warned that tightening U.S. restrictions could hurt overseas growth. The United States has represented as much as 20% of its sales during reported periods.

Blocking new Chinese models could give American robotics companies breathing room. It could keep low cost foreign machines from securing early market share before domestic competitors reach volume production. But it could also deprive those same U.S. companies of affordable development platforms and Chinese made components.

Protection Carries A Price

The big question is whether a broad country based restriction will make the American robotics sector stronger. American startups often build on global supply chains. Motors, gearboxes, batteries, sensors and electronic assemblies may come from China even when the software and final machine originate in the United States. Cutting off finished Chinese robots may help domestic manufacturers, but cutting off key components too quickly could raise their costs.

The United States wants to build more data centers, solar farms and battery projects to meet rising electricity demand from AI computing. Yet, restricting import and use of Chinese power electronics may raise prices or delay projects when replacement suppliers lack enough production capacity. Robot restrictions could deny American users access to inexpensive machines while China continues testing them in its own huge domestic market.

Even if these import blocks may reduce exposure to remote disruption, it may impose new challenges and costs for the already overheated tech sector. China has already warned that it may retaliate against the robot and inverter measures. Beijing’s Commerce Ministry accused Washington of misusing national security and discriminating against Chinese companies.

A cycle of restrictions could split the robotics market into competing technology blocs. American firms may sell inside the United States and allied markets. Chinese firms may expand through Asia, Africa, Latin America and the Middle East. Separate technical standards, software systems and component networks could follow.

The current U.S. position treats robots and power electronics as part of the same strategic stack as advanced chips. That stack begins with electricity and computing. It extends through models, sensors, actuators and communications networks. At the far end sits a machine that can perceive its surroundings and act without constant human control.

America retains deep strengths in AI research, semiconductor design, software and venture capital. China holds a significant advantage in industrial scale, component production and factory deployment. Without domestic manufacturing capacity, skilled workers, reliable component suppliers and aggressive commercial adoption, a wall around the U.S. market could protect an industry that remains too small to compete abroad. The FCC’s restrictions may buy time for American robot makers, but time will tell if this helps or hinders the U.S. market’s ability to compete locally and abroad.