Last month in Indonesia, the presidents of the United States and China met to discuss the question of Taiwan, the war in Ukraine, and the issue of human rights in the People’s Republic. The biggest recent development between their two countries, however, might have been something much lower profile: In October, the U.S. Commerce Department introduced a comprehensive ban on technology exports to China, including cutting-edge semiconductor chips—the foundational components for advanced computers and other electronic devices—as well as software for designing them and equipment for manufacturing them. The new ban covers not only American companies but any company in the world that uses American semiconductor technology. And it forbids all U.S. citizens, residents, and green-card holders from working for Chinese chip makers. The threat to China’s high-tech industries—whether in military applications, artificial intelligence, or the surveillance technology Beijing relies on to monitor its citizens—appears so grave that Gregory Allen, of the Washington-based Center for Strategic and International Studies, called the new U.S. approach to China’s tech sector “strangling with an intent to kill.” How deadly is this?
Chris Miller is an assistant professor of international history at Tufts University’s Fletcher School of Law and Diplomacy, and the author of the new book Chip War: The Fight for the World’s Most Critical Technology. In the defining great-power competition of the early 21st century, Miller says, U.S. President Joe Biden has made an unusually aggressive move to halt technological progress in China’s military by exploiting the West’s critical advantages in the chip sector. The strategy could end up doing real damage to China’s economy over the long term; but in the meantime, it could also spur Chinese President Xi Jinping to create an independent Chinese production chain—and all the uncertainties for the U.S. that would come with it.
Michael Bluhm: What’s behind these new export controls?
Chris Miller: Over the past two decades, the military balance between the U.S. and China has shifted dramatically in Beijing’s favor. Over the next decade, China is likely to continue developing its military capabilities and shifting the balance even further—for instance, by building more ships and airplanes than the U.S. will. And China is likely to continue producing more military systems than the U.S. will—and to deploy more of these systems in contested regions across Asia, above all in the Taiwan Strait.
All of this has been happening with the U.S. becoming increasingly concerned about China’s intentions—in Taiwan and elsewhere in the region. Beijing’s moves in the South China Sea, in particular, in recent years suggest that, where it can, it’ll use its military power to bolster its political influence and territorial control.
Now, given that the Chinese government has a standing policy to establish territorial control over Taiwan, there’s been a growing concern in the U.S. that the shift in the military balance between it and China could destabilize the region—by opening the door for Beijing to try to take Taiwan or otherwise transform Asia’s security landscape.
The question for the U.S. is what to do about this. One option would be to give up on trying to maintain the status quo in Asia—but there’s a bipartisan consensus in the U.S. that this isn’t a good idea. So the Americans have to find some way to halt the deterioration of their military position relative to China’s.
Bluhm: How does the U.S. administration expect the export ban will address the problem?
Miller: Because no one expects any major increase in the U.S. defense budget—to build a lot more ships, planes, and missiles, for example—the administration has to address the problem qualitatively, by creating better systems than China’s.
Because China is the manufacturing workshop of the world, the solution can’t be in physically manufacturing better systems. But the Americans and their allies have a qualitative lead in computing and with the semiconductors that make computing possible. That’s what the U.S. administration targeting with these controls: They’re trying to keep, and ideally grow, the U.S. advantage over China with advanced computing and sensor capabilities in military systems, which already rely on a lot of semiconductors. And they’re trying to reduce China’s access to the advanced data centers and supercomputers used to design or train the AI systems deployed in military devices.
There’s been a growing concern in the U.S. that the shift in the military balance between it and China could destabilize the region—by opening the door for Beijing to try to take Taiwan or otherwise transform Asia’s security landscape.
At its simplest, the U.S. objective is to reduce China’s access to cutting-edge computing, to retain America’s advantage in it, and then to deploy this advantage to military and intelligence systems.
Bluhm: You mention implications for military competition, AI, and big data. How big of a deal are these?
Miller: It is a very big deal. The controls are written so that they won’t only affect the Chinese government or military; they’ll affect all of China. Restricting the transfer of certain advanced chips and chip-making equipment into the country altogether, these controls are going to have a real impact on civilian firms in China that have no direct relationship with their government.
The U.S. wrote the controls this way because they concluded that their previous effort—which targeted only the transfer of equipment to the Chinese military—hadn’t worked. The U.S. had no control of chips once they entered China. So even if the chips weren’t supposed to go to a Chinese military facility, many companies in China openly advertised that they could use U.S. chips and technology for the Chinese military.
There will be collateral damage to Chinese tech firms. There’s no doubt about that. The U.S. administration was aware of this in writing the new controls, but they decided it was a price worth paying for putting more effective limitations on the Chinese military and intelligence apparatus.
It’s now going to be more difficult to build an advanced data center in China than it will be anywhere else in the world. And over time, this has to mean significant constraints on Chinese tech firms’ access to high-performance computing capabilities.
Chinese civilian firms can still access advanced data centers abroad. But they can’t undertake these advanced computing capabilities in China—unless China finds a way around the new controls.
There will be collateral damage to Chinese tech firms. There’s no doubt about that. The U.S. administration was aware of this in writing the new controls, but they decided it was a price worth paying for putting more effective limitations on the Chinese military and intelligence apparatus.
Bluhm: You talk about the effects inside China. But the new export controls could also have important geopolitical consequences. The relationship between Washington and Beijing seems likely to define a lot of global politics in this century. How do you see the new controls affecting this relationship?
Miller: The U.S. recognizes that there are zero-sum aspects to U.S.-China competition, and the new ban is a zero-sum move. For a long time, it’s been popular to say, We compete in certain spheres and cooperate in others. And Biden’s people themselves still embrace that framing in some respects. But if you look at this administration’s first two years, it hasn’t actually found many ways to cooperate. It’s focused largely on competition.
Of course, some competition can be positive; it can make everyone better off. But this competition over chips isn’t designed to make everyone better off; it’s designed to hold China back as the U.S. moves forward. That’s not unusual, as such; it’s standard practice among great powers, in the midst of a rivalry like this one, to try to prevent their adversary from getting access to technology. But it does add to the recognition that the competition between the U.S. and China has moved to a more zero-sum phase than it was in just a couple of years ago.
Military technology and intelligence capabilities are where competing countries’ relationships are most likely to become zero-sum. And if the U.S. goal is to preserve and develop the American advantage in defense-intelligence capabilities, it’s almost inevitable that the U.S. administration will be thinking not only about helping America advance but also about holding China back.
Bluhm: Xi Jinping has long made leadership in high-tech industries a core element of his vision for China’s future. How is the Chinese leadership responding here?
Miller: As you say, the Chinese government has prioritized leadership in high-tech for some time. But the Chinese government has also been aware, for at least a decade, of its reliance on imported hardware. Since 2014, Xi and the Chinese government have been focused on finding ways to reduce China’s reliance on imported chips and chip-making technology.
Government programs out of Beijing have supported this agenda at the national, provincial, and local levels. They’ve made some strides here, too, though not as many as they’d have liked. But where things stood in 2014 is largely where they stand today. China is still heavily reliant on imported technology from countries like the U.S., Japan, and Taiwan—which are geopolitical competitors and adversaries. This isn’t a comfortable position to be in if you want to be a global leader in technology.
Some competition can be positive; it can make everyone better off. But this competition over chips isn’t designed to make everyone better off; it’s designed to hold China back as the U.S. moves forward.
In some ways, this Chinese desire for technological self-sufficiency—or at least independence from the U.S.—has been there longer than the U.S. has prioritized semiconductor chips. China started prioritizing them in 2014, whereas it wasn’t until 2016 or 2017 that U.S. policymakers became fully aware of how strategically important they were for defense and intelligence applications.
These controls won’t meaningfully change China’s chip policies because China was already fully focused on reducing chip imports and becoming less reliant on U.S. technology to produce chips at home.
Bluhm: Biden and Xi met on the sidelines of the recent G20 conference in Indonesia. Would they have discussed these export controls?
Miller: There’s no public information about that, and the Chinese government hasn’t really responded to the controls at all, apart from condemning them rhetorically.
Bluhm: How are Chinese high-tech industries responding?
Miller: Chinese tech firms are responding in two different ways. One is to try getting around the controls by using different chips. Some companies have begun offering slightly less-capable chips that are exempt from the U.S. ban. This is within the ban’s letter of the law, as it were, and also the spirit: The Americans are perfectly happy for chips that aren’t on the technological cutting edge, and don’t introduce new advancements, to be sold to China.
The other way Chinese tech firms are responding is to try building up Chinese chip-making capabilities. In recent years, China has had some meaningful success in the chip-design part of the process, but in chip manufacturing, it remains very heavily reliant on imported tools, especially from the United States.
So this is going to be a long-term process; it’ll take China at least half a decade to make any real progress. But Chinese corporations will be increasingly focused on it.
Bluhm: How much damage do you expect this could do to China’s high-tech industries—and its broader economy?
Miller: In the short run, the economic impact will likely be limited, because the controls only affect a few kinds of chips—and only a couple of companies that sell chip-making tools. So it’s not going to have a big macroeconomic impact in the first year or two, outside the semiconductor space.
Neither is it likely to trigger a collapse of investment in artificial intelligence. Some decline is likely, but the government in Beijing will step in where it has to—and if you’re a Chinese firm selling to the Chinese market, all your competitors are facing the same restrictions and will still be trying to develop better AI algorithms and utilize high-performance computing capabilities. There’s still a lot of incentive for Chinese firms to focus on AI.
But if the U.S. controls are effective in preventing China from accessing the types of advanced chips they’ll need for AI and data centers in the long run, that’ll have a much bigger impact. Chinese firms will be unable to access chips that’ll be widely accessible outside of China. And the economic impact of that will only grow over time.