The last time Chinese leader Xi Jinping entered the Oval Office, the US-China economic relationship looked very different. In 2015, the notion of tariffing more than two-thirds of Chinese imports would have been hard to imagine. A decade later—and mere days before Xi returns to the White House to meet US President Donald Trump—it’s the geoeconomic reality we live in.
The September 24 summit comes against the backdrop of an ongoing trade truce, with an apparent understanding that any new US tariffs will not exceed an effective rate of 30 percent. Yet the two sides still face a host of other challenges—from artificial intelligence (AI) governance to headwinds in Beijing’s own economy. Our experts break down the economic issues hanging over the Trump-Xi meeting—and what comes next.
Click to jump to an expert analysis:
Jessie Yin: Cooperation on AI will be constrained by the US-China tech race
Beth Baltzan: Washington still lacks a coherent strategy for competing with Beijing
Jeremy Mark: Domestic woes might force Xi to look west
Cooperation on AI will be constrained by the US-China tech race
Last weekend, leading AI companies in the United States proposed slowing down frontier development over safety concerns and called on their Chinese counterparts to do the same. But just hours later, Trump dismissed such concerns as “hoaxes” and emphasized the stakes of AI development for the US-China tech race: “because whoever wins AI wins.”
This is emblematic of the fundamental contradiction facing Washington and Beijing: As leading AI developers, both countries should have an interest in cooperating on global AI governance. Yet they are locked in cutthroat competition. US Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng discussed AI in their talks over the weekend, agreeing to continue the dialogue. But the technological rivalry leaves little room for consensus on guardrails, especially as US companies continue to raise national security concerns about Chinese AI labs conducting industrial-scale distillation of US frontier models.
The United States has used economic tools to restrict China’s access to advanced semiconductors and slow its tech development. But AI distillation, which uses outputs from a stronger AI model to train a smaller one, presents a different challenge. The United States could use the entities list, export controls, and the foreign direct product rule to restrict Chinese labs’ access to leading US models. But the harder question is how to distinguish legitimate AI training techniques from illegitimate attacks.
Beijing, for its part, might decide that the benefits of continued AI development outweigh the risks. For China, AI is a critical chance to reduce its dependence on the United States and gain an edge in a technology that could shape the global economy and security landscape for decades to come. The Chinese government has worked to extend its strict regulatory regime for the internet to AI and has begun drafting a mandatory national standard for AI agent safety. While not dismissive of AI’s potential dangers, Chinese policymakers have historically approached AI as a technology that can be managed through technical standards, security assessments, and state oversight. Whether or not this approach can sustainably mitigate risk, Beijing’s willingness to push forward with AI development raises the stakes for containing distillation of US models, especially if leading US AI companies agree to slow frontier-model development.
The AI discussion this weekend largely focused on setting the agenda for a US-China AI dialogue mechanism that would cover everything from AI safety to economic competition. The US side also proposed a new AI safety notification mechanism for Trump and Xi to consider at their meeting. But the heated rhetoric surrounding US-China AI competition—from heads of state to AI engineers—means that anything resembling a formal agreement is still far off. For now, AI innovation will continue to move faster than cooperation on how to govern it.
—Jessie Yin is an associate director at the Atlantic Council’s GeoEconomics Center.
Washington still lacks a coherent strategy for competing with Beijing
The United States will not arrive at the summit in a position of strength. The Trump administration has overplayed its hand in the trade war with China: it imposed and then lifted export controls on chips that Beijing nevertheless declined to buy, and launched a war with Iran that has exposed vulnerabilities in the US military while driving up energy prices. Both countries would benefit from easing tensions while grappling with the very real problems that afflict their relationship. But it’s not clear whether this summit can put them on that path.
The United States and China have fundamentally different economic models that in many ways are simply not interoperable. World Trade Organization rules, designed in the heyday of the Washington Consensus, did not contemplate a world in which members would remain non-market economies. The consequences are felt by workers all over the world: labor compensation’s share of all economic inputs is declining not only in countries such as the United States and Germany, but in China, too. And recent talk of letting Chinese manufacturers of electric vehicles into the US market suggests that the Trump administration may not appreciate this reality.
The administration claims that cooperation on AI safety is a summit priority. At the same time, fearmongering about China is becoming the military-industrial complex’s currency of choice—particularly in the technology sector, where AI is being framed as an arms race that the United States must “win,” with “American” frontier labs treated as national champions. This martial tenor is not only at odds with the administration’s stated summit goals; it also undermines efforts to make genuine progress toward guardrails or governance. More broadly, this disconnect reflects the lack of a coherent US approach to economic competition with Beijing, which will continue to complicate efforts to manage fundamental incompatibilities between the two economies.
—Beth Baltzan is a nonresident senior fellow at the Atlantic Council’s GeoEconomics Center.
As the US focuses on limiting economic exposure, don’t expect a major breakthrough on trade
The first Trump administration spent months negotiating with China over a new trading relationship. Those talks produced the “Phase One” agreement. However, China ultimately failed to comply with many of the agreement’s key commitments. The Biden administration kept—and expanded—the tariffs imposed on China for its alleged “theft of US intellectual property and trade secrets.” Then COVID-19 added a new bipartisan concern: the vulnerability of critical supply chains. Eight years later, Washington’s approach has slowly but fundamentally shifted. US policymakers have largely stopped trying to persuade Beijing to change its economic policies and are instead focused on limiting the US economy’s exposure to its market-distorting practices.
From the Trump administration’s perspective, those efforts have been relatively successful. The United States has the strongest economy in the Group of Seven, China has so far been unable to damage key US industries such as autos and steel, and Washington is working to develop alternative sources for critical minerals and other goods where China’s state-supported industries could create supply shortages. That leaves little incentive for Washington to reintroduce economic reform as a centerpiece of talks with Beijing. At the same time, Xi and his team appear to be doubling down on exports to support China’s economy. Of course, they would welcome greater access to the US market, but it is hard to see what China could offer in return.
Trump’s recent suggestion that he would be open to Chinese automakers building cars in the United States could give Xi another opening to seek market access. But any such investment would undoubtedly face strong opposition from China hawks on Capitol Hill, and Trump would likely demand more than Xi is willing to give. That points toward a narrow trade agenda at this week’s meeting, rather than a major breakthrough. The “Board of Trade” idea that emerged from the last Trump-Xi summit could be a vehicle for cooperation, but with little enthusiasm on either side for broader economic coordination, it will likely remain focused on practical issues such as soybean sales and critical minerals.
—Stephen Vaughn is a nonresident senior fellow with the Atlantic Council’s GeoEconomics Center.
Domestic woes might force Xi to look west
Xi comes to Washington facing unprecedented economic headwinds at home. While he and his minions will project their usual hubris about China’s achievements, there should be no mistaking that Xi needs more US orders for Chinese goods just as much as Trump seeks increased sales of American products.
Exports have been propping up the Chinese economy ever since the government’s COVID-related shutdowns and its measures to rein in a property market bubble undercut business and consumer confidence, leaving domestic demand in the doldrums. In August alone, property-related investment fell 20 percent, car sales plummeted 24 percent, and loans to businesses were down by half from 2023 levels. Youth unemployment hit 17.9 percent in July, including many of the millions of recent university graduates, and gig workers now represent nearly half of the country’s workforce. The Rhodium Group estimates that cumulative Chinese economic growth from 2021 to 2025 amounted to about 2 percent, a far cry from Beijing’s claim of average annual growth of 5.4 percent over those years.
The industrial production that powers China’s overseas sales—an increasing proportion of which is high-tech goods tied to the current AI boom—is the economy’s one bright spot. Chinese exports to the United States fell 27 percent in 2025 from 2024, largely because of Trump administration tariffs. Although they have been rebounding rapidly in recent months—up 34 percent year on year in August—the United States now accounts for a much smaller share of those exports than in the past. Last month, it bought about 10 percent of Chinese goods compared with a peak of 21.6 percent in 2017. At a moment when every cargo container counts, Xi will be seeking an agreement that reopens the door to more of his country’s industrial goods at lower tariffs.
—Jeremy Mark is a nonresident senior fellow with the Atlantic Council’s GeoEconomics Center. He previously worked for the International Monetary Fund and the Asian Wall Street Journal.
Related Reading
Image: US President Donald Trump at a bilateral meeting with Chinese leader Xi Jinping in Beijing, May 14, 2026. Source: REUTERS/Evan Vucci.
