How high-end semiconductor chips threaten US-China relations : The Indicator from Planet Money A week before the Communist Party of China's 20th congress kicked off, the US government announced comprehensive new rules blocking China's access to advanced semiconductor chips. It's a sign the Biden Administration can't disentangle security concerns from trade with China.

For sponsor-free episodes of The Indicator from Planet Money, subscribe to Planet Money+ via Apple Podcasts or at

Why the US cut China off from advanced chips

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript




The main news coming out of China over the last week has been the Communist Party of China's 20th conference, where President Xi Jinping is likely to secure his third term.


PRESIDENT XI JINPING: (Non-English language spoken).


But behind the scenes, something else has been happening, something huge. On October 7, the U.S. announced sweeping export controls on advanced semiconductor chip technology to China - you know, those chips used in supercomputers and even in some self-driving cars. Greg Allen is a technology and security expert.

GREG ALLEN: This policy overturns 20-plus years of U.S. trade policy and national security policy towards China. It's an incredibly big deal.


WOODS: And I'm Darian Woods. Today on the show, a tectonic shift as the U.S. bans advanced semiconductor chips for China. We break down why these chips matter, what the bans entail, how we got here and what this means for China, the U.S. and the world.


WOODS: To appreciate just how big this ban on advanced chips is, just look back at the year 2000. China had been slowly opening up its market since the late 1970s, and it was now a considerable player in manufacturing. The U.S. voted to accept China into the World Trade Organization, which regulates trade.


BILL CLINTON: Good afternoon.

WOODS: Here's President Bill Clinton at the time. He's speaking at the White House Rose Garden.


CLINTON: We will be exporting, however, more than our products. By this agreement, we will also export more of one of our most cherished values - economic freedom.

WONG: Were we all just a lot happier in the '90s, in the early 2000s?

WOODS: It's a level of earnestness that I don't know if exists any longer.

WONG: Oh, totally. But you know, this was the mainstream thinking at the time. That's according to Greg Allen. He's the director of the AI Governance Project at the Center for Strategic and International Studies.

ALLEN: The theory was that as long as we're in globally integrated trade networks, we don't need to worry too much about what technologies we transfer to China. We just need to make sure that we maintain our leading position, and we can do that by being open to free markets.

WOODS: But then came 2012, a year where two pivotal moments in world history happened. One was Xi Jinping becoming China's leader. And the other was a little geekier. It happened at an online competition hosted by Stanford and Princeton. It was the ImageNet Large Scale Visual Recognition Challenge.

WONG: The task was to build a computer system - like, some kind of artificial intelligence - that would draw from a massive database of over a million images and figure out what that image was. Like, was this an oven or a bowling ball or a dog? And if it's a dog, what dog breed is it?

WOODS: It's a English terrier.

WONG: (Laughter).

WOODS: And a team of computer scientists based in Canada called AlexNet turned up for the competition using a system called machine learning. And they'd taken the system to the next level using graphics processing units, GPUs, the kind of graphics chips you use for gaming.

WONG: Graphics chips are really good at simultaneously running the same calculation many, many times, like identifying aspects of hundreds of thousands of images.

WOODS: Like looking at hundreds of thousands of pictures of ovens.

WONG: Is this an Easy-Bake oven? So AlexNet's accuracy swept the floor.

ALLEN: And that is really the modern artificial intelligence revolution. It sort of got started in computer vision but quickly went on to speech recognition, language translation and just a host of other applications where machine learning does quite, quite better than traditional software.

WOODS: Greg says machine learning is now used everywhere in the military, from guided missiles to analyzing vast quantities of satellite imagery.

ALLEN: So rather than looking, you know, at every single glacier or every single picture of empty ocean with nothing in it, they pre-process all that collected imagery using artificial intelligence. And the artificial intelligence will come up with answers like this airfield didn't have very many planes in it yesterday, but today there are 20 new aircraft, which I think have an 85% chance of being fighter aircraft. I recommend a human analyst take a closer look at these images.

WONG: And to train the artificial intelligence systems, you need advanced computer chips. And so when advanced computer chips give your military an edge in almost everything where software is used, they become even more prized for security.

WOODS: And that brings us to where the China-U.S. relationship around semiconductor chips has gotten incredibly tense. Now, a lot has happened over the last decade. For instance, the Chinese government has launched cyberattacks on U.S. companies to steal trade secrets. Now, that is not great, obviously, if you're a U.S. chip company like Intel or NVIDIA.

WONG: And Greg highlights a key moment in 2017 when the Chinese government passes its national intelligence law. This law said the Chinese government could compel Chinese citizens to hand over technological know-how in matters of national intelligence.

ALLEN: And so what this means is any day, if you are, for example, a Chinese citizen who is working for an American company at their overseas in China subsidiary, the Chinese government might pay you a visit someday and say it is now your duty as a citizen of China to steal the technological secrets of your American employer and bring them back and give them to us. And by the way, this is the law. So if you refuse to comply with us, we can jail you for this.

WOODS: And Greg says it became clear that when it comes to dealing with China, separating trade from security concerns was impossible. So the Biden administration has decided that because the U.S. cannot guarantee that its advance chips will not be used for military purposes in China, they don't get any at all.

ALLEN: It is now a China-wide restriction. This new policy just fundamentally changes what it's like to do technology, research, development and manufacturing in China.

WONG: It also stops China from making its own high-end computer chips because American companies are barred from sending equipment and software to China that's used to design and manufacture these advanced chips. This is an area that the U.S. dominates.

ALLEN: China has climbed very high in the semiconductor industry, but it was built on a foundation of U.S. technology. When the current equipment breaks, there will be no spare parts, and there will be no advice from the American manufacturer on how to make do with what you've got.

WOODS: In fact, dozens of American executives working on advanced chips in China may have to either quit or apply for a special license, which is really uncertain, or give up their ties to the U.S. entirely. Multinational semiconductor companies are already transferring their American stuff away from anything to do with China.

WONG: Greg has one major caveat, though. China can still smuggle in parts from other countries. And he says this is a huge task for the agency charged with enforcing the ban. But still, he says the screws have tightened across large swaths of the Chinese semiconductor industry.

ALLEN: This policy is designed to not only slow that, this is designed to reverse that.

WOODS: I mean, this is a profound point. This is the end of U.S. optimism over the possibility of China as a trading partner in the world where one rival's advances spurs the next and we both kind of go into the future together. It's a real shift. It's a loss.

ALLEN: And it's not just a shift for the United States. You know, you've seen actions out of the Chinese government like the crushing of Hong Kong, like the atrocities in Xinjiang. And all of that has sort of led the entire world to reassess the desirability of a strategy based on peace through trade, when it is clear that these authoritarian regimes see the world quite differently and that what - the effects that we thought we were going to have in pursuing that policy over the past two decades have just not come true.

WONG: Yeah. Russia's invasion of Ukraine does seem to have called that theory into question.

WOODS: It really has. And it throws us into this uncertain world. Like, I'm thinking, how will China respond to all this? You know, all eyes in Beijing are on the party congress right now. But given the scale of the chip ban, it seems hard to imagine we won't be hearing from Xi Jinping, assuming he gets that next term, right?

WONG: Does he have a challenger?

WOODS: Not that I'm aware of.


WONG: This show is produced by Nicky Ouellet with engineering from Gilly Moon. It was fact-checked by Dylan Sloan. Viet Le is our senior producer. Kate Concannon edits the show. And THE INDICATOR is a production of NPR.

Copyright © 2022 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.