Consider This from NPR - What does global leadership on AI look like?

Episode Date: September 20, 2026

President Trump and Xi Jinping will meet in Washington, D.C. this week as both sides look to gain an advantage in artificial intelligence development and face a reckoning on the risks of AI. In this moment, what does global leadership in AI look like?Support public media with NPR+ and enjoy perks for over 25 podcasts like this one. This show’s perks include bonus episodes and sponsor-free listening. Learn more at plus.npr.org.This episode was produced by Daniel Ofman and Henry Larson. Our director is Jason Fuller.It was edited by Sarah Robbins.Our interim executive producer is Courtney Dorning.See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.NPR Privacy Policy

It's consider this, where every day we go deep on one big news story. Two of the most powerful people in the world will meet in Washington this week. President Trump and Chinese President Xi Jinping last convened in Beijing in May, where they walked a red carpet, were greeted by children and military bands, but there was little progress on policy. Now, the two are expected to talk about Iran, tariffs, and artificial incentives. I think there's no more important topic that the U.S. and China should be talking about than AI. That's Sarah Beran, a former U.S. diplomat who worked on China issues at the National Security Council.

Both countries are facing a sudden reckoning on AI. In America, a growing number of industry insiders are raising concerns that the technology is dangerous. President Trump maintains that beating China and AI is paramount and has tried to tamp down mounting unease, telling attendees at a tech summit a few days ago, not to worry. The robots will not be taken over. The AI will not be taken over the rest of the world. The whole thing is a hoax. He's also just pledged to create a, quote,

AI force. Consider this. China and the United States are the centers of global power. And as both sides look to gain an advantage in artificial intelligence development, what does global leadership in AI actually look like? From NPR, I'm a lady. Sanadwarni. Let's consider this from NPR. At any meeting between two world powers, each side

comes with a list of wants. And for China and the United States, those lists encompass some of the most important issues of our era, trade, national security, and the status of Taiwan. Also near the top of that agenda will be technology and the fierce competition between Washington and Beijing on issues ranging from the production of microchips to artificial intelligence. I called up Ann Newberger. She's a former U.S. Deputy National Security Advisor for cyber and emerging technologies. I wanted to learn more about those priorities. And I started by asking her what she will be focused on for this week's summit.

The U.S. and China are locked in a race for AI supremacy and cyber dominance. And neither country is going to sign a broad treaty that frees tech development. President Trump needs to send a clear message that disruptive cyber intrusion into civilian infrastructure crosses a red line. So establishing, for example, dedicated bilateral hotlines for AI-enabled incidents is a pragmatic step to manage any of that kind of miscalculation. You just said disruptive cyber intrusions. Now, you wrote an article in Foreign Affairs magazine in which you describe something you call, quote, digital choke points. How do they play into the U.S.-China relationship? China has one of the largest and most aggressive.

offensive cyber programs in the world. They focus not only on espionage, but also on pre-positioning in critical parts of countries' infrastructure, power, pipelines, water in the U.S. and in U.S. allies and adversaries of China, countries like Japan, Singapore, the Philippines, Taiwan, and Australia. And when we describe choke points, much as we've seen Iran, the Houthis, work to try. to weaponize geographic parts of the world where they are in a unique position because they have potentially some degree of control over a geographic space. We have a concern that China will use its prepositioning in digital infrastructure to disrupt the operations of those infrastructure in a period of crisis or conflict. You also talk about certain deficiencies that the U.S. has

that leaves it vulnerable if China decides to take advantage of them. What are some of these vulnerabilities and what can be done to address them? We're a large country and we have thousands of water systems around the country. We have thousands of electricity systems around the country. Many were built in the last 30, 40, 50 years before the age of cyber warfare. So given the size of U.S. infrastructure, given the degree as people call it of tech debt, of old systems that really are hard to update, and given the fact that they're largely owned by private companies who until recently didn't have any obligations to maintain a set of

cyber standards in terms of the security of those systems, that's a real vulnerability for the United States. So tell me about how AI impacts this dynamic. AI is a game changer. It's a dual-use technology. It's a game changer for attackers. But I think it's an even bigger game changer for defenders because AI can take a lot of information and help quickly analyze it. There's a massive amount of data coming off a network related to just the operations of a system, think a national power system, and AI can help gather all of that information, identify what may be anomalous activity, unusual activity, which is often the first sign of malicious cyberactivity on a network. And I actually believe it helps defenders even more than attackers, because attackers always

had an edge. An attacker only needs to find, you know, one digital door in. A defender has to be watching every door and window. And in the age of AI, attackers can jiggle every door and up continuously all the time. So as a result, what AI can do for defenders in terms of running on a network, gathering all the data, finding something that's anomalous, or analyzing old code to generate patches for that code is a game changer because it makes that work possible in a way that really wasn't when we only were able to do this through people's hard work and skill. What should President Trump be communicating to Xi Jinping about this technological competition between Beijing and Washington? You know, Washington and Beijing can try to agree on one area of

shared risk, preventing AI models from offering actionable assistance in generating chemical or biological weapons. That is one area that potentially the countries could work together. In fact, over the last number of years, one of the positive areas of U.S.-Chinese cooperation was in countering the spread of fentanyl. Much of the precursor chemicals are produced in China, and as you know, it was really wreaking havoc in American cities. And the U.S. and Beijing, U.S. and China, worked together to put in place some controls, and it was a positive area of cooperation.

So cooperation is possible when there are joint areas. of interest. And I think that preventing AI models from offering that kind of assistance in generating chemical or biological weapons is in both countries' interest and could be a foundation for some kind of work together. I wonder, within the U.S. government, what's the most misunderstood aspect of China's use of cutting-edge technology? I think the big area of debate has been in the area of whether leadership means having the best models or whether leadership means the world running on a technology as its foundation. In the U.S., the approach has really been trying to ensure that U.S. models are always at the cutting edge. China, in contrast, has a playbook,

which we've seen in other technologies like 5G, of recognizing that good enough, lowest cost technology is what many countries and companies around the world will pick. Indeed, when we look at Silicon Valley, we see a lot of companies, even in the United States, building on Chinese open source models. So I think the biggest difference to your point is perhaps a recognition that we need both. It is important that the world reflect democratic ideals in the way the technology is applied and because the technology has so much promise. So coming up with ways to help help more companies around them and countries around the world build on American models, recognizing that cost is often the factor, should be a good part of the American strategy for AI.

And I think that's one of the reasons that we've seen more U.S. companies start building open source models because they recognize that for many countries and companies around the world, the appeal of open source and that it is much lower cost is real. Anne Newberger is a former U.S. Deputy National Security Advisor for cyber and emerging technologies. Thank you so much for being with us. Really a pleasure to be with you, Alyssa. This episode was produced by Daniel Offman and Henry Larson. Our director is Jason Fuller.

It was edited by Sarah Robbins. Our interim executive producer is Courtney Dorney. It's Consider This from NPR. I'm Alyssa Adwarnie.