Most of us think of AI as a chatbot on a screen, but the technology is moving into the physical world. Governments are now treating both the brains of these systems and the robot bodies that house them as critical national assets, leading to a new wave of trade rules that could reshape how technology is built.
The US government has issued a ban on many foreign-made robots, including human-shaped units, citing concerns that these machines could gather sensitive data in homes or workplaces. At the same time, American policymakers are debating whether to block Chinese AI models that are made freely available for others to download and adapt. These moves are part of an effort to protect the domestic AI and robotics industry from being undercut by cheaper, rapidly improving competition from China. Reports suggest such a ban on Chinese models could prevent businesses from realizing an estimated $25 billion in annual savings.
The cost of closing the door
To understand why these bans are so contentious, it helps to know how researchers develop new AI. Most AI is built by training a model—a massive mathematical engine—on enormous amounts of data. During this process, the model learns to identify patterns, essentially creating a set of internal settings called weights that determine how it processes information. When a company releases these weights, it is called an open-weight release. This allows outside developers to take the finished engine and customize it for their own needs without having to spend the millions of dollars required to train it from scratch.
In robotics, the situation is similar. Academic labs and startups often use inexpensive, pre-built robots from companies like the Chinese manufacturer Unitree to experiment with new software. Because a domestic robot might cost nearly 60 times more than a comparable imported one, these cheap machines are the workhorses of university research. By banning these imports, the government aims to force the industry to rely on US-made parts, but researchers warn that this could inadvertently stop the very innovation the US is trying to protect, as labs struggle to afford the tools they need to advance the field.
The tension here is between security and speed. Proponents of these bans argue that foreign tech acts like a trojan horse, capable of spying on our most private spaces. However, the rapid rise of Chinese AI models that match the performance of American giants like OpenAI and Anthropic complicates the strategy. If the US restricts access to foreign software and hardware, it might keep data safer, but it could also create an isolated bubble. The challenge for the future is whether we can keep our systems secure without cutting off the collaborative, open approach that has historically allowed technology to evolve so quickly.
Liked this one? The next lands at breakfast.
Every story in tomorrow's AI news, rebuilt in plain English — five minutes, sources linked, free forever.
By joining you agree to receive Article's daily newsletter — unsubscribe in one click. Privacy