CSIS Press Briefing: Previewing the Trump-Xi Summit

This transcript is from a CSIS Press Briefing hosted on September 17, 2026.

Nina Prieur: Hello, everyone. Welcome to the CSIS press briefing, “Previewing the Upcoming Trump-Xi Summit.” I’m Nina Prieur, communications director here at CSIS. And I’m joined today by CSIS experts who will share their insights into what to look for going into the summit, what we can expect to come out of the summit, and more. 

Just a couple housekeeping notes before we get started. Each of our speakers will give several minutes of introductory remarks, after which we’ll turn to your questions. If you want to ask a question, please type it in the Q&A window or raise your hand to ask it verbally. We’ll be distributing a transcript of today’s call shortly after its conclusion, and the transcript will also be made available on CSIS.org. With that, let’s go ahead and get started. 

I’ll turn first to Edgard Kagan, senior advisor and Freeman chair in China studies at CSIS. Edgard, over to you. 

Edgard D. Kagan: Good morning, everybody. And thank you very much for joining. And thanks to our CSIS External Relations team for organizing this.

Obviously, this is the return visit from President Trump’s visit to Beijing in May. This is quite significant, because it is – for one, it’s the first visit by Xi Jinping to Washington since his state visit in 2015. And it’s also significant in that it comes so soon after the president’s visit in May. This is something that is very much out of the norm in U.S.-China relations, to have two back-to-back state visits in such a short period of time. 

I think one of the things that is sort of striking is that, in many ways, the visit is the message. There’s been a surprising lack of signaling by either the White House or the PRC on major goals for the visit. I think that the way that they are framing it is very much, this is about managing the relationship, which they very much believe can only be done by regular engagements at the leader level. And I think that both President Trump and Chairman Xi share the view that each of them is the only one who can manage the other, and that the only way to manage the bilateral relationship is by their interactions. 

I think the major elements that I will be looking for, and I’ve got much smarter colleagues on the line who can fill in many of the details, but one is an extension of the tariff truce, which is essentially limitations on new U.S. tariffs in return for continued flow of rare earths and critical minerals. I think the key thing is that this was done in Busan. It was one year. I think the expectation is the two leaders will commit to renewing this and extending it. I think the Chinese believe that there will be an announcement of around $30 billion of duty-free or certainly low-tariff imports from China, as a return for the commitments they made for purchases in May. I think the White House believes that they will also be getting some additional purchase commitments. 

I think something that has become obviously very big in the last few weeks, but was already on the docket going back to May, is the question of AI. And my colleague Aalok Mehta will be able to go into more detail on this. I think that at one point my expectation was that there would be some sort of meeting and some sort of announcement on a framework for AI cooperation. Personally, I expected it to be rather devoid of substance. I think in light of the brouhaha in the last few weeks and some of the Chinese statements, I’m more skeptical that that will happen, but it’s certainly a possibility and it would be significant if it did. 

Another big issue, obviously, is Taiwan. The Chinese clearly want a continuation of the freeze on arms sales, and I think that they will push for that, though I think that the real significance of what happened in May was less the arms sales themselves than the fact that the president came out essentially mouthing Xi Jinping’s framework on Taiwan.

And so I think the Chinese will be looking to see if they get anything more on that. I think that they will be stressing the importance of continued U.S. restraint with regard to Taiwan, and I think that this is something that’s being watched very closely by U.S. partners and allies in the region and, obviously, in Taiwan. 

Another thing that has developed I think in the last few weeks is the question of some detained American citizens, in particular, the declaration that there – that a U.S. researcher named Min Xin, who was detained in June, has been unlawfully detained, and on top of Chen Youlin, both of whom are researchers who are essentially being detained for things having to do with research, supposedly, as best we know. And so I think there’s an expectation the president will push for their release. 

I think that the other big issue, obviously, that is going to be discussed will be Iran. The president is, clearly, very, very focused on Iran, and I think that the key will be the degree to which he’s willing to make specific asks and the degree of receptivity from the Chinese side. 

One of the things that is very interesting is China appears to be increasing its purchases of oil on the global market and that’s something that could have an impact on prices, and so I think the president will be very focused on figuring out how China’s concerns about its supplies can lead to China being willing to use some of its leverage with Iran to push for a settlement. 

Things that I will be watching are – this is a state visit, and state visits have a certain protocol. There are certain standard elements to them. One of the things that has evolved over the years is that while there’s the formal day of the state visit, which will be the 24th, most of the time, particularly when leaders come from Asia or Europe, they arrive earlier to help adjust with jet lag, and there’s usually a dinner or some sort of event with the couples, the president and the first lady and the visiting leader and their spouse. And so it’ll be interesting to see whether or not this happens. 

This is not a formal element of U.S. protocol, but it has evolved that way over the years, and so if that does happen I think that will be a sign that the president is very focused on signaling warmth and a close personal connection with Xi. So I think that that’s something that will signal sort of the degree of warmth and the positive vibes surrounding the visit.

Also, I will be watching very closely to see if either leader talks about a return visit to either Shenzhen for the APEC summit in November or the G-20 summit in Miami in December. And, you know, as for those of you who’ve been involved with these kinds of visits, planning them is really, really hard. The logistics of moving a leader across the Pacific is very extensive and very demanding, and so I suspect that certainly at the working levels in both the White House and Zhongnanhai that people will not be heartbroken if these visits don’t happen.

And so it’s possible that they’ll signal affirmatively that they will. It’s possible that they’ll refer to it. I think that would be quite significant as a sign that there will be further exchanges, which keeps the pressure on both systems to try and avoid actions which could disrupt the detente and this effort for constructive strategic stability that came out of the May discussions but really are, and I would argue, a through line from the Busan meeting. 

So with that, let me pass it over to my colleagues, who know far more, and I would just stress that you’ve got to take this with a grain of salt. I think that we’re all trying very, very hard to figure out what exactly both sides or which direction they’re going to, and so I think that inevitably we will be surprised in some areas.

Let me turn it now over to my colleagues. 

Dr. Prieur: Thanks very much, Edgard. 

Next, we have Scott Kennedy, senior advisor and trustee chair in China business and economics at CSIS. Scott, go ahead.

Scott Kennedy: Thanks, Nina. And good to have everybody on the call this morning. 

Just very briefly, I want to say something about the likely outcomes, a little bit about the nature of the relationship, and what I think is a lost opportunity in the current diplomacy that is ongoing and will unfold over the next week. I would agree. I don’t want to go down through the entire list, but I agree with Edgard’s list of likely things. Number one is whether or not they’ll extend the ceasefire. The Chinese have wanted an agreement to extend the ceasefire through the end of the Trump administration. The Trump administration just wants to extend it a few months to maintain leverage on China. I think they’ll end up somewhere in the middle, probably at about a year.

I think we’ll also see deals related to agriculture, aircraft sales, et cetera. And these may take the form of signing ceremonies. We’ll see them talk about AI. I think also they will – I know the president would like to have there be an announcement about the next two visits, which I think the U.S. sees as both intimately tied to each other – the president going to APEC in Shenzhen with Xi Jinping visiting Miami for the G-20. It’ll be interesting – I don’t know if they’ll discuss it here and publicly, but will President Putin be at both, including in Miami? 

I think to me, one of the most interesting parts of this – of next week will be the Chinese business delegation. We still don’t have a clear sense of who will make up that delegation, other than some rumors about perhaps BYD’s leader and a couple other companies. But I’m not sure what role they will have, but these are some impressive companies that have made a lot of progress in the last few years. And if they are given any stage to share their views and contribute to the discussions of the summit itself or on the sidelines, I think that will be worth watching and perhaps one of the only real unclear parts about what’s going to be done. I don’t see a lot happening on Taiwan, Iran, Russia, other than discussions without significant progress. 

I do think, to pivot to my second point, is that this is the framework for the engagement. Yes, both sides are trying to keep things calm. It’s in both sides’ interests. But I think this is largely mostly to the benefit of China. I think now this is Xi Jinping’s world and we’re all living in it. The administration’s China strategy, I think, was poorly designed from the beginning. They thought they could use massive unilateral pressure to force China to make concessions. And that did not occur. In fact, the opposite did. It ended up being the U.S. that backed down. We’re not implementing, clearly, a revolution of the global trading system and security system in a way that has made China back up. 

The administration’s red lines on China that we’ve typically had related to industrial policy, Taiwan, South China Sea, human rights, that is hard to find in the administration’s approach. Instead, they want an avoidance of war. They want to keep markets up, and anything that makes President Trump look good. And so very different kinds of red lines that the U.S. now has than in the past. I think also it’s important to note that the disorganization within the administration, the lack of a clear policy process, has made it very difficult to make these summits have a lot of deliverables, which is why the two sides have to meet this coming weekend to really iron out much of what is going to be discussed and agreed upon. 

I think this is a – my final point – is that this is real lost opportunity. China, yes, it is doing amazing things in technology across a whole variety of industries, and we’ll see that in the composition of the business delegation, but China’s got a bunch of challenges. It has – it’s growing very, very slowly, not 5 percent or anywhere close to it. It’s got a massive debt problem, unemployment is growing not just for young people, but for others. There’s massive – there’s underemployment problems. Much of the world is really unhappy with being hit with this wave – tidal wave of Chinese exports that are a product of overcapacity. You see a lot of purges going on inside the Chinese political system and a lot of popular anxiety.

So China is facing a lot of domestic challenges, but the U.S. is basically giving Xi Jinping a pass. We’re not constraining the Chinese, really, in any significant ways, and the U.S. could. We could have a much more effective approach, one that would also involve collaboration with allies, which we are alienating with many of our other policies.

So I think Edgard is right. This is going to be a summit with – that’s relatively underwhelming with a small number of deliverables. But I think that works mainly to China’s advantage. Let me stop there.

Dr. Prieur: Thanks very much, Scott.

Next we have Bonny Lin, director of the China Power Project and senior adviser at CSIS. Bonny, go ahead.

Bonny Lin: Thank you, Nina. And thank you, everyone, for joining us.

And I just want to say that I agree with many of the points that Edgard and Scott have mentioned already. So I’m not going to repeat them, and I’ll try to focus a little bit more on a couple of security/defense-related issues.

But before I get to that, I do want to make two points at the top. One is, to Edgard and Scott’s discussion about the planning for the Trump-Xi summit, I think from the U.S. side there is a sense of no more than reciprocity. In other words, I don’t think Xi Jinping is going to get anything more than what Trump got in Beijing. So I think that will limit in – that will limit what Xi may or may not get in D.C.

The other high-level point I want to mention is, Edgard talked about the constructive relationship of strategic stability. I think for the Chinese side, what they are really looking for after the summit is to be able to fill out the details under this slogan. I think it’s understandable from their perspective and from the global perspective, if they’re making the argument that this is now defining the U.S.-China relationship, that maybe after the May meeting it was not well-formulated and there weren’t that many details. But to have another meeting between the two leaders and still come out with no details underneath this slogan I think will probably be viewed as a disappointment from the Chinese side. So I would be looking for at the highest levels what each side says falls under the constructive relationship of strategic stability, and particularly if either side defines any particular elements as falling underneath that.

But now let me transition to some security- and defense-related issues. So I agree very much with what Edgard has already mentioned on Taiwan. I don’t think Xi’s goals on Taiwan has changed. And I think, if anything, there might be even more optimism from Beijing’s end that meetings between Xi and Trump can help manage U.S. policies on Taiwan.

So, again, we expect Xi to bring up Taiwan. And we also expect Xi to try to, if possible, push the United States to delay or stop U.S. arms sales to Taiwan. And there’s also expectation that having all these meetings – not only this one, but potential APEC and G-20 meeting – would further delay any Taiwan presidential transits of the United States. As you all probably – as you all probably know, previously, prior to 2025, Taiwan presidents transited the United States on average of one to two times per year. It hasn’t happened in almost two years now.

One trend I do want to point out is, if you recall prior to the meeting between the two leaders, there were periods of significant decreases in PLA air activity around Taiwan and there was various speculation that it was a Chinese attempt to show goodwill towards the United States. And some of us even heard that directly from our Chinese interlocutors. I don’t believe that that was the reason the air activities decreased, but, again, this is not the trend we are seeing now. Since the May Trump-Xi meeting, we’ve seen air – PLA air activities increase compared to before May. And beyond the air activities, PLA naval vessel activities against Taiwan has largely remained constant in the past two years. But particularly since May, we’ve seen Chinese official ships or Chinese coast guard vessels increase operations around Taiwan. And they’ve particularly increased operations to the east of Taiwan, where they have been engaging in patterns of monthlong law enforcement patrols. So this is a relatively new development.

In terms of the U.S.-China military-to-military relations, we saw Admiral Paparo from U.S. Pacific Command meet with the PLA Eastern Theater commander, Yang Zhibin, on August 31st on the sidelines of the Indo-Pacific chiefs defense conference in Vancouver, Canada. That was the first time we’ve had meetings at the theater level, military-to-military contacts, since 2024.

We also saw at the Shangri-La – sorry, at the Xiangshan Forum, that is occurring right now in Beijing, that Chinese Defense Minister Dong Jun was less pointed in his remarks about the United States and U.S. allies, and explicitly did not mention Taiwan in his speech – which in many ways echoed Secretary Hegseth’s remarks at the Shangri-La Dialogue. I do think it’s possible, if they are looking for items on the military security side to fill out what may be under the constructive relationship of strategic stability, that adding more military-to-military engagements could be on the cards. 

In terms of Iran, Edgard has already talked about this a bit. I would just note that China has been very cautious in engaging Iran. Edgard mentioned that China has been purchasing more oil from the global markets. And that reflects a decrease in Chinese purchases from Iran because of the tensions and because of the difficulty of getting oil out of the Strait of Hormuz. I think it’s important to note that President Xi was very cautious in the very brief pull-aside he had with the Iranian president during the SCO meeting. In fact, Chinese outlets didn’t even mention the meeting. So also, when the Iranian foreign minister met with Wang Yi this week – actually, yesterday in Beijing – we didn’t see him have an opportunity meet with Xi Jinping. 

Let me touch very briefly on Russia and North Korea, and mainly in the context of the Xiangshan Forum, and what that could be teeing up for, particularly the APEC meeting. Scott had mentioned that there is a possibility floating around that perhaps Russia would be at the APEC meeting. And that could provide an opportunity for President Trump to meet with Putin there. Russia’s deputy minister of defense said at the Xiangshan Forum that Putin will go to China in November. And the Kremlin has repeatedly suggested a trilateral between U.S., China, Russia at APEC. And China has not denied this. And this came up again at the Xiangshan Forum. So I think that’s very interesting to watch. 

Also, at the Xiangshan Forum this week, we saw the first high-level North Korean delegation to the Xiangshan Forum since 2019. And this was led by North Korea’s Vice Defense Minister Kim Kang-il, who’s also a prominent voice on North Korean security and defense policy. At the Xiangshan Forum he explicitly said that the U.S.-led military buildup and nuclear alliance against – nuclear alliance of Washington, Seoul, and Tokyo justified North Korea’s nuclear expansion. The fact that he was using the Xiangshan Forum to make these remarks provides further evidence of China’s de facto acceptance of North Korea’s nuclear status. 

But what’s super fascinating and not reported much in Western media, and printed in Chinese language media, is he also said, according to Chinese media, during the Xiangshan Forum, that the North Korean military will abide by the guidance of the two leaders and strengthen strategic communication exchanges and cooperations with the PLA. If you recall, when Xi traveled to Pyongyang in June, the first time in seven years, he suggested to Kim Jong-un that the two sides need to deepen military engagements and cooperation. So this is North Korea answering back and saying, yes, we are doing – we’re taking steps to do so. 

And I raise this in the context of this summit and APEC because there’s been lots of rumors of the possibility that APEC could also be a venue for China to host Kim Jong-un, in China. So President Trump could meet with Kim in China. And I think China would be quite interested in having more of a role in facilitating U.S.-North Korea communications. There’s also, of course, the possibility that after APEC or before APEC President Trump could meet with Kim. But either way, China is shoring up its influence with the anticipation that there may be a Trump-Kim meeting. 

I’ll wrap it up here. Thank you. 

Dr. Prieur: Thank you, Bonny.

Next, we have Aalok Mehta, director of the Wadhwani AI Center at CSIS. Aalok, the floor is yours. 

Aalok Mehta: Hi. Thank you. And it’s nice chatting with all of you. 

So this summit may feel like it’s coming very quickly in terms of quickly following on from the May summit. But in AI time, it feels like it’s been several decades. And there are two, I think, notable developments that are important to keep in mind for framing how the U.S. and China might talk about AI at the summit. So the first is this set of containment incidents, where AI agents from multiple companies were able to escape out of their sandbox and hack websites on the internet. We’re continuing to get details about how extensive this was, and also about how sophisticated the communication between agents was, as they sort of shared information and collaborated in engaging in some of these hacks. 

And the second, which happened very recently over the course of the past two weeks, is that issues around sort of existential or catastrophic risk from AI have really exploded into the mainstream, first through the resignation of an Anthropic researcher. And his resignation post has now been seen more than 170 million times on X. He resigned out of concerns about the pace of AI development and the potential risks that that poses to people and the economy. And then, following on that, within a few days Anthropic CEO Dario Amodei issued sort of a proposal or a call to “pace the frontier.” And he also announced sort of a unilateral action in which Anthropic is going to invite a third-party organization to embed within the company and sort of conduct oversight over how it approaches model development.

And I think these are a mixed bag for what we might expect coming out of the summit in terms of AI. So the cyber incidents, I think, are the really important framing set of incidents here. We know that this has spooked American policymakers. And it builds on concerns that they’ve had ever since the release of Mythos back in February, and the increasingly powerful cyber capabilities of AI models. And that we also know that Chinese policymakers are also thinking about these issues, and that they are also considering whether there needs to be more government intervention to give governments more ability to know what’s going on inside the labs, and maybe to provide more time for society and for policymakers to respond to these developments. 

On the other hand, some of the rhetoric that we’ve seen following the “pacing the frontier” proposal has complicated the possibility discussions at the summit. In particular, some of the discussion in that essay, you know, sort of contained statements that Chinese officials are interpreting as fearmongering or sort of raising particular pointed concerns about Chinese AI development. And so that is a little bit of a complicating factor. I do think that overall there are a lot of incentives for the U.S. and China to have a substantive discussion on AI in this – at the summit. 

And I think the most likely way that that would proceed is that the two countries might come and sort of discuss a narrow set of issues related to, first, definitions of what it means to develop AI safely, sort of basic safety practices. Second, setting up some sort of communication hotline that would allow the U.S. and China to be in touch around emerging or developing AI issues. Third, something more specific around cyber incidents. And I think this is likely the most concrete action that could come out of the summit. And it would be something like the U.S. government agrees to do more monitoring of cyber incidents caused by U.S. companies, the Chinese government agrees to do more monitor monitoring of incidents related to Chinese companies, and then there’s some sort of information sharing process where each government redacts sensitive information and then is able to send that to the other government. 

And finally, all of this will only work if there are some sort of verification mechanisms in place. And so I think that there will be some discussion about both technical and governance tools that could allow information to be shared without having – necessarily relying just on sort of verbal statements from officials. And then the final thing I’ll say is that, given the pace at which AI is developing, I do think it is very likely that these discussions will continue, but they will not sort of continue on the summit track; that there will be more technical discussions happening at a faster pace and more often as U.S. and Chinese policymakers try to grapple with some of the – some of the big emerging issues from the rapid development in AI technology. 

Dr. Prieur: Thanks very much, Aalok. 

Let me quickly mention that after our next expert we’re going to be turning to your questions. So if you’d like to ask a question, please go ahead and type it in the Q&A window, or as we begin our discussion you can always raise your hand to ask it verbally. 

Next, we have Michaela Simoneau, fellow for global health security with the Global Health Policy Center at CSIS. Michaela, please go ahead.

Michaela Simoneau: Thanks, Nina, and thanks to the whole CSIS external relations team, and good morning, everyone. 

I am on this call because I recently returned from a trip to Beijing and Shanghai with a group of CSIS scholars where we were focused on bilateral engagement on health and the bioeconomy, and during our trip we met with academic – (off mic) – U.S. government officials, and Chinese government officials who are working across health, foreign affairs, and counternarcotics issues, and our argument for making such visits has been simple.

Despite the deep skepticism and division that we have between our two governments, we believe we need to engage with one another on health security matters. There’s areas where cooperation on health could yield mutual benefits on the shared issues our populations are facing such as cancer or mental health or rare diseases. 

But we’re also engaged in intense competition over dominance in biotechnology, and without regular high-level dialogue, misunderstandings and miscalculations become much more likely.

We heard from a lot of Chinese partners that they’re worried about that risk, and we should be, too. But unfortunately, it’s unlikely that the September 24th summit will include much discussion or really any deliverables on health, biotechnology, or the bioeconomy, and that’s a real missed opportunity.

It tracks, however, with recent summits between the two presidents where health stayed in the background. In Busan last fall, the president struck a deal for the U.S. to ease tariffs in exchange for greater Chinese action to control fentanyl precursors, and that was framed, largely, as a public health effort to reduce U.S. overdose deaths.

This past May in Beijing, health issues were even more absent. There was only one biotech executive who joined President Trump’s delegation, there were no major biotech deals announced, and the U.S. didn't mention health at all in its readout.

But despite the lack of specific health deliverables, the May meeting was an important signal for the U.S. and especially the Chinese system. 

As some of my colleagues have mentioned here, there’s this new construct of constructive strategic stability that we heard from nearly every Chinese official we met on our trip, and this may provide some sort of basis for future engagement on health and health security beyond the summit later this month. I’ve been watching three urgent areas where we could be engaging more strategically. 

First, we need to enable continued innovation while managing the risks of our commercial interdependence in the bioeconomy. China has been investing heavily in biotech and it’s nearing parity in many areas with the United States, and there are several commercial incentives driving this including Chinese government subsidies, industrial power, and fast, low-cost trials. They’re driving massive amounts of U.S. capital in licensing deals to China, despite increasingly harsh Chinese restrictions on data and capital flows. 

And there’s two sides of this. The rapid innovation could bring major benefits to American patients, but the transition in the bioeconomy and U.S. dependence on China in several domains stokes real anxieties around whether U.S. industry could be at risk of coercion or domination by the Chinese and whether U.S. innovative capacity could be hollowed out like other industries. 

The House Select Committee certainly thinks these two issues are real and has focused on controls to try to slow Chinese growth. But the calculus in the bioeconomy isn’t quite as straightforward as with critical minerals or microchips. This is a highly diverse sector and it requires several different sorts of policy tools. 

The National Security Commission on Emerging Biotechnology and the House Biotech Caucus have emphasized the need for a complementary industrial strategy with guardrails and incentives that boost our U.S. competitiveness, and this would involve investments in data systems, in talent, and in regulatory reforms here at home.

The second area is we should be cooperating in specific health areas where the exchange of policy and practice could yield important mutual gains. On our trip, we heard wide agreement that we could cooperate on areas such as accelerating innovation on lifesaving medical discoveries, supporting an aging populace in both our countries that are facing a growing burden of chronic ailments, containing emerging infectious disease threats, and coordinating diplomatically to strengthen international institutions.

And the third point I’ll mention is as the world’s lead superpowers we have a responsibility to engage on some of the most sensitive health security challenges. We’ve seen this work. The U.S. and Chinese law enforcement professionals have cooperated closely and really effectively on fentanyl by scheduling precursor chemicals, exchanging fugitives, and facilitating leader-level exchanges among law enforcement officials. And these efforts are part of the reason we’ve seen a 50 percent drop in fentanyl-related deaths between 2023 and 2025 in the U.S.

We have a responsibility to carry forward similar sorts of cooperation on AI safety, especially guardrails against the use of AI to develop bioweapons. Both large language models and AI-assisted bio-design tools make it easier for bad actors to design, to obtain, and eventually to deploy bioweapons. And Anthropic’s report last week detailing how it blocked seemingly state-sponsored efforts to build bioweapons should be a serious alarm bell on that front.

As other scholars have mentioned, this is an extraordinary year. We could have up to four face-to-face meetings between the presidents. And biotech, especially health issues, might get more play later this fall at the APEC summit in November, regardless of whether or not President Trump makes an appearance. There are dedicated working groups on regulatory harmonization and health that have met for months in preparation for that summit, including U.S. officials.

But on balance in health and the bioeconomy there’s this overwhelming pressure to compete, not to engage. We deeply distrust each other’s motives. Rhetorical commitments on public health over the past several years and several administrations of summits haven’t always been matched by action. And we’re still stuck in this unresolved debate over COVID origins. But we face really great mutual threats, so it’s in our national security interest to reprioritize dialogue on these issues of health, economic, and national security.

I’ll stop there. Thanks, Nina.

Dr. Prieur: Thank you, Michaela. And thank you to all of our speakers.

We’re at that time now where we’ll go ahead and open up the floor for questions.

(Gives queuing instructions.)

And we’ll get started here with a question from Katrin Büchenbacher from Germany. The question is: The constructive relationship of strategic stability defined in May has been described as a truce or a timeout. Do you think it will be kept alive or even become more practical during the upcoming state visit? How do you think both the U.S. and China are using this relative calm in bilateral relations to their advantage? And who do you think benefits more? Edgard, would you like to kick us off with your thoughts on that question? Then we can open it up to others.

Mr. Kagan: Sure. It’s a great question.

I think one of the things that’s very interesting – and I want to preface this by saying I haven’t been in a meeting with Chairman Xi since 2022 and I haven’t been in meeting with the president since 2025. So, you know, things have changed, and so I don’t want to – I don’t want to oversell my own knowledge.

But my sense is that when the two of them are together – when you see Xi Jinping and Donald Trump together, each of them believes that time is on their side. The president very much believes that over time the U.S. economy is going to grow, the U.S. will get stronger, the U.S. will keep making advances in AI and on technology, and that time – and also, the U.S. will develop alternative sources of critical minerals and rare earths, and so that time is on the U.S. side. I think when Xi Jinping sits down with Donald Trump, he very much believes the U.S. is in hurtling decline and that time is on China’s side.

I think where their interests overlap is for a variety of reasons, particularly linked to domestic economy. Neither of them has an interest in – at least at this point in an escalation on trade issues. I think that – I agree completely with Scott. I think the Chinese response to the tariffs that the administration put on is a very powerful signal that they are not going to be pushed around, and I think they’ve been very effective in using that. I think at the same time, as Scott said, the Chinese economy is growing very slowly. And it is quite possible, I would argue even likely, that the real growth rates of the U.S. economy is perhaps even outgrowing China, certainly growing at comparable rates, which for China is a significant slowdown. So I think that Xi wants time and space. He wants a trade truce, to the extent that he can have it, so that he can focus on his domestic economy.

I would just note that the real question is going to be – and this speaks to what Aalok was discussing – how this plays out on technology, how this plays out on energy, how this plays out on the development of semiconductor capabilities. And in all these areas, there is a great deal of friction and I think that the Chinese very much want more space. They want a lifting – and their supporters in the U.S. as well want a lifting – of some of the restrictions on exports, particularly on semiconductors. And so all these things will play into this.

I would just make one other comment, that going into these summits traditionally, certainly when I was involved, you wanted to try and have momentum by showing that the U.S. relationships with key allies and partners were strong; and that this is something that, you know, the Chinese paid close attention to, and the Chinese tried to do the same. What is quite striking is if you look at the last six months, and certainly since – in the last five months since Trump went to Beijing, you see a lot of disarray in U.S. alliances and partnerships. And I would just highlight the G-20 meeting, finance ministers meeting in Asheville, really highlighted some deep divisions between the U.S. and some of our traditional allies and partners. Also, the friction with Canada and the escalation of the trade conflict, but also Canada seeking associate membership from the EU, that’s significant. The decision to unilaterally freeze and reduce U.S.-ROK exercises in an attempt to curry favor with Kim Jong-un also has caused friction with the ROK. And I think that Japan, even though they’re too polite to say this, is frustrated that the U.S. hasn’t come out more strongly in their support, given the pressure China’s putting on. And the Southeast Asian countries were deeply disappointed and deeply concerned by both Secretary Hegseth’s remarks at the Shangri-La summit in June and then the speech that Undersecretary Bridge Colby gave in Manila, where he managed to not talk at all about the South China Sea or the pressure China’s putting on the Philippines.

So I think all these things need to be set against the two recent trips Xi Jinping has done to – for the Shanghai Cooperation Organization and also the BRICS, which have really highlighted that China has strong relationships with a growing number of key partners. So I think this really does give China a certain amount of momentum, and undermines the president, and makes the U.S. look weaker than it should.

Dr. Kennedy: Hey, this is Scott Kennedy. Just to provide a little bit more to this very good question, and very simply, the definition of whether or not this constructive strategic stability is still holding is whether or not the two leaders are meeting right now. That’s the definition.

The U.S. and China are both doing a lot of things bilaterally with regard to the other that still is about pressure. The U.S. has taken a lot of actions over the past few months that China’s unhappy with, including most prominently with regard to the FCC actions. The Chinese have also done a lot of things with regard to extraterritorial rules and other things. We’re not calling that a breaking away from this because they’re meeting next week. If they didn’t meet next week, then that would be a sign that that relationship that they defined in May is breaking down.

That doesn’t necessarily apply to the potential meetings in Shenzhen and Miami later this year. But for right now, the fact that they’re meeting, that’s the measure of whether or not we are in this period of constructive strategic stability.

Dr. Prieur: Thanks very much.

Our next question is from Chris Buckley with The New York Times. Chris asks you, Bonny, to: Please tell us more about the state of discussion about Taiwan and arms sales to Taiwan in Washington, and especially inside the Trump administration. In particular, what factors are likely to weigh on if, when, and how Trump announces more arms sales to Taiwan?

Dr. Lin: Thank you, Nina. And thank you, Chris, for the question.

So my understanding is the Trump administration still is maintaining close and low-profile contacts with the Taiwan government. The relationship is still very close. Particularly on the defense side we’re closer than ever.

In terms of the arms sales, I think there is some urgency to – well, I think we’re not going to meet it. I think there’s some urgency on Taipei’s end as well as some of the defense contractors involved in the proposed 14 billion sales to be able to get the announcement before end of September, but I don’t think that’s going to happen. My understanding is the most likely scenario is that the package is going to be broken up. And they’re discussing right now how to break up that package. But the question is the timing of the announcement of those packages – those smaller packages. 

My worry, and that’s why I was talking about the potential for President Trump to meet Vladimir Putin at APEC. Also a potential that the APEC meeting could provide a forum or a way onto President Trump to meet with Kim Jong-un. I think, in addition to what may or may not be achieved in this September summit, I think China is dangling a number of carrots in front of President Trump in order to discourage President Trump to not provide arm sales, at least before the APEC meeting. And what I worry about is I really don’t see, given the close timing between this summit and the APEC meeting, that there is any window of delivery of the arms sales, at least before the APEC meeting. 

What I been hearing is if President Trump goes to the APEC meeting, it’s very likely Xi Jinping is going to come to the G-20 meeting, which then pushes the calendar for the next smaller arms sales delivery to basically the end of this year, if not next year. And the question is, will there be more meetings that pop up? To Scott’s point about how, and Edgard’s point, about how important it is for the two leaders to meet, I think, in many ways if the two leaders keep on meeting at such a consistent, and regular, frequent schedule, it’s going to make arms sales to Taiwan very difficult, even smaller packages.

Dr. Prieur: Thanks, Bonny.

I’m going to combine two questions into one here, with you in mind, Scott, for this next one. Eleanor Mueller from Semafor asks: Curious if the experts could talk about sanctions on Iran and whether they expect Xi and Trump to discuss the possibility of sanctioning Chinese banks to pressure Iran, which Monday’s Treasury announcement seemed to dance around.

Followed by a similar question from Ben Marks: In the context of Operation Economic Outcast, how hard do you expect the U.S. to press China on severing financial support to Iran during the summit? 

Dr. Kennedy: Both very good questions. And it is amazing how quiet the administration has been since the announcement of Operation Outcast concerning China’s commercial relationship with Iran. And I think that’s an indication that maintaining a stable relationship with China is trumping, pardon the pun, U.S. policies on Iran, in this regard. And I don’t see an increase in willingness to assume more risk vis-à-vis China on this because of China’s leverage that they have effectively used. And so I expect China to use the same kind of language that it’s used on Iran over the last few months again, and but without a breakdown – even if the U.S. is unsatisfied, I don’t think that is going to lead to an escalation between the U.S. and China. 

I think actually this is something that Edgard, I think, might be able to speak to better than myself. 

Dr. Lin: Actually, if I could jump in here, Scott, really quickly before Edgard speaks. I do think there has been slightly more U.S. pressure against China via Iran. We’ve seen that the Treasury has targeted Russia’s VTB bank. And VTB is Russia’s second-largest bank, and the only Russian bank with a full branch inside China. We know that Iranian leaders have talked about how their traders are getting paid through Russian bank accounts in China. So I think right now the administration is still careful. But, again, Secretary Bessent is going to talk with his counterpart this weekend. And he has explicitly noted that Iran sanctions will be on the agenda. And I don’t think we’re going to get any of those sanctions before the Trump-Xi meeting, but I wouldn’t rule out the possibility that more action could be taken against China. And I don’t know, Edgard, if you agree or not. 

Mr. Kagan: I agree. I think both of you hit it on the head. I would just say that going back 20 years to North Korea-related issues, our friends at the Treasury Department have always been deeply, deeply reluctant to go after top-tier Chinese banks, and there’s a whole bunch of reasons having to do with the international financial system, their concerns about possible retaliation, other issues. I find it hard to believe that now we’re going to suddenly really change our approach.

At the same time, there are quite a few things that can be done to ratchet up pressure by going after intermediaries. There’s always the possibility of going after smaller Chinese entities, and so I think that the real question is going to be how much the president and how much Xi Jinping are seeking to maintain as positive a relationship as possible.

And, you know, the president is playing a significantly weaker hand with regard to Iran than he was even in May when you can argue it wasn’t great. And at the same time, China is increasingly worried, at least in my view, about the ramifications of continued closure of the Strait of Hormuz, both on its own economy but, more importantly, on the global economy on which it increasingly depends as a destination for exports. 

So I think that what you’ll see is an avoidance of direct major confrontation on high-profile issues, but there will be a lot of maneuvering in the trenches, and I think that the Chinese have consistently shown they don’t want to cross U.S. red lines, and I think this also gets to the question of intelligence sharing and whether or not the president will go after them on that. 

As I think people on the call know, this is something that Presidents Bush and Obama raised with their PRC counterparts and that their administrations put a lot of pressure on China on when it looked like China was sharing equipment and designs and exporting components that were used for very deadly IEDs that targeted U.S. troops in Iraq. So it’ll be interesting to see if the president, given that there seems to be some information suggesting Chinese intelligence was involved in an attack that led to U.S. deaths, the president will do the same. 

Dr. Prieur: Thanks. Thanks, all.

We have a question here from Gloria Teal from Fox. Gloria says that she missed the first 15 minutes, so if this has already been addressed we can ignore it, but I don’t think we’ve addressed it. She says: Over the past few days, there’s been rumors that Xi suffered a health event at the BRICS summit and is wondering if any of the experts have insights on this. Is it possible the summit may be delayed? Any insights?

Dr. Lin: I guess I’ll just open to say we periodically see these rumors about Xi suffering health issues, but I haven’t seen anything saying that the summit will be delayed. I don’t know, Edgard or Scott and others, if you’ve seen anything.

Mr. Kagan: Look, I take these kinds of things with a grain of salt, and even – I mean, the source for a lot of this is Indian press reporting and even some of the Indian press reporting notes that he appears to have continued to conduct some activities after the supposed stroke. 

You know, the kind – issues surrounding the health of leaders are always very, very sensitive. The Chinese are exceptionally good at controlling information flow on this. I personally am skeptical only because we’ve seen rumors like this going – you know, pop up fairly regularly and they don’t seem to have really been validated. You never know. 

I think the proof will be if the summit is delayed, if suddenly the Chinese announce or the U.S. announces that it’s going to be put off, that would, in my mind, be a pretty strong data point suggesting there really was a health issue. At the same time, we haven’t seen anything suggesting that, and if anything, we continue to see exactly the kinds of preparations and discussion that we would expect surrounding a meeting on the 24th of September. So my sense is it’s probably going forward and this is probably not correct, but obviously we’ll see what comes out in the coming days. 

Dr. Prieur: Thanks very much. 

We’re just about to hit time here, but I think we have time for one more question. I just want to say in advance that I’m sorry to those whose questions we couldn’t get to. We’re seeing fantastic attendance and participation on this call. So please do reach out via email or phone. We’ll be happy to connect you with our experts who will be engaging in all sorts of ways in the week to come and beyond that.

For our final question we’ll turn to Jessica Gardner, who asks: In any discussion around AI safety frameworks, what are the top priorities that Xi and Trump bring to the conversation? And how do you expect any tension there might play out? Certainly in the AI realm, Aalok, I imagine you’ll have some insights, but our other scholars as well. Who wants to take this one?

Dr. Mehta: Yeah, I’m happy to start. So I think the most important sort of guiding principle here is that the U.S. and China won’t agree to anything that sort of doesn’t allow the countries to continue to compete on the merits of the technology. They still want to be able to sell their respective tech stacks abroad. And that leaves sort of a very narrow set of issues that they can have productive conversations on. 

Another tension here that they’re dealing with is the sort of claims that a lot of Chinese technology is distilled off of U.S. technology. So these claims, as well as the possibility of sort of ratcheting up export controls on semiconductors, both sort of would put China in a defensive mindset and make it harder for them to agree to anything. A third factor here is that the U.S. still has the most advanced models in the world. And so I do think that there is a way in which China feels like the U.S. has to take more responsibility because not only is their technology farther ahead, it is likely that it – that there are unreleased models that the frontier companies have that are even more powerful than what we know about publicly.

That being said, you know, just in the last 24 hours we’ve seen Bessent say that there is – that there is interest in talking about shared risks. And I do think the most likely areas of discussion are around shared risks, shared approaches to addressing those risks that sort of have roughly equivalent impacts on Chinese and U.S. labs, information sharing arrangements, and then verification mechanisms related to knowing both what datacenters are being used for and verification mechanisms for information that is being shared between the U.S. and Chinese governments.

Mr. Kagan: : Look, I agree completely with what Aalok said, but I think that there’s a – there’s also the fact that China is deeply suspicious of anything that looks like it’s designed to lock in U.S. advantages. Also, there is a fundamental distrust between officials in both countries that’s going to make it really hard to have the kind of information sharing that I think would be necessary to have some confidence that both sides were implementing whatever they agree to do. 

And I think that a lot of people compare this to strategic arms talks between the U.S. and the Soviet Union in the sixties and seventies. I think the big difference is that there was – there was more understanding of the technology, which was, frankly, moving less fast, even though at the time it looked like it was moving fast. And also, there was some sense of what each side’s interests are that made some level of confidence surrounding verification possible. There were always doubts. I mean, both sides always worried the other side was cheating. I think now we don’t even have that basic level of trust on AI. And so I think it will be very, very hard to come up with a mechanism that is anything more than symbolic right now.

I know Scott probably has some views, and they’re probably better than mine.

Dr. Kennedy: I think that was well said, both of you. So I’m happy to leave it at that – the questions – the answer to that question there. 

Ms. Simoneau: I would just very briefly add – this is Michaela – China and the United States also, their leading AI companies are taking very different approaches to the guardrails they have built into their models, or whether they’re open weight. And that generates very different sorts of risks, especially when you’re talking about catastrophic biological risks, and especially when you’re talking about these bespoke, highly technical models that are trained on specific sorts of data. So there’s a bit of talking past one another when we’re discussing the fundamentals of the technology. 

Dr. Prieur: Thanks very much.

It looks like we are just about at time here. So I want to thank our speakers and all our participants for joining us for this call. And, again, please do reach out if we didn’t get your question, you want to make sure to engage, or we can be a resource to you on this issue in any other way. We’re here to help. And as I mentioned at the top of the call, we’ll have a transcript distributed to you. And it will also be available at CSIS.org. So with that, I hope everyone has a good rest of the day. And thanks again for joining us.