Analysis

The CCP faces a conundrum: how does the party maintain “resilient” growth amid the necessary disruption of pivoting away from China’s largest trading partner, the United States? China’s economy is a political weapon in the CCP’s arsenal, and the party will deploy all means necessary to extract the most from the US in the short term while pivoting to alternative markets in the medium-to-long run. A pressing question for CCP policymakers will therefore be how the current US and China trade dispute catalyzes the party’s approach to global trade. Chinese outbound investments to the US, for example, have been paused since April while negotiations with Washington continue, and are down a whopping 95% since their peak in 2016. Yet while bilateral investment flows are likely to restart after a trade deal is reached, they are unlikely to rebound to earlier levels. 

For China, investing in the US has always been a tradeoff between economic (and knowledge) gain versus national security; now, that tradeoff is shifting significantly in favor of the latter. Self-sufficiency is more than a buzzword in China, it is fundamental to how the CCP is shielding itself from the US. Moreover, increasing trade with other economies, particularly in BRICS and the Global South, helps China insulate itself from US export controls, sanctions, or other economic levers that Washington may try to use as bargaining tools. As we argue below in “Business Matters,” this view of economies and trade relations as extensions of state power is increasingly the norm not only in China, but also in the US and elsewhere.

If resilience is measured in terms of security, then maintaining growth is calculated not just as economic expansion, but importantly also how insulated new economic activity is from geopolitics. This means that corporate decisions that take multiple years, like infrastructure investments, will have to adapt to increasingly fast-paced geopolitical changes. It also means that companies engaged in trade will increasingly need to factor geopolitical calculations into their own risk analyses and balance sheets. 

Global trade flows are already starting to follow this shift in prioritization: We are seeing a global-scale “strategic interdependence,” prompted by pressure from Washington and Beijing, for countries and companies to increasingly choose one side over the other. Crucially for businesses, this means that corporate decision making will increasingly need to weigh traditional cost-benefit calculations alongside equally important alternatives, including national security risks, potential exposure to cyber risks, and how to respond to employees (particularly c-suite executives) caught in geopolitical crossfire. Playing one side off against the other, as Nvidia recently found out, will likely be an increasingly risky–though not impossible–strategy. US firms will therefore need to be significantly more selective about their China-based investments and engagements than in the past. Failure to do so could leave companies exposed to financial or political punishment.

Analysis

While some US companies are wising up to how to handle China’s increasingly brazen attacks against foreign corporations, they are also increasingly facing pressures from the US government. Two related stories demonstrate how politics is shaping business in the US: the US government’s pressure on Intel’s CEO to resign and Apple’s announced investment in facilities across the US. Both highlight how even if companies are not interested in politics, politics is very much interested in them, particularly when those companies have ties to China. 

First, Intel’s CEO, Malaysia-born Lip-Bu Tan, came under fire for his alleged ties to China, before being lauded as an exemplar after meeting directly with President Trump. The attention on Tan allegedly appears to concern the US government’s desire for Intel to spin off or sell its fabs and spin off Intel Foundry to TSMC, a move backed by Intel board chairman Frank Yeary but opposed by Tan. US government pressure against Tan, seemingly appeased by Tan’s meeting with President Trump and promise for the US government to take a stake in the company, highlights the precarious political landscape that some company executives, particularly in the tech industry, may face. 

Second, Apple’s announced investments in the US also speak to the political moment, and appear particularly targeted towards the US government’s priorities. Apple announced that it would invest $600 billion–up from $500 billion announced in February–and emphasized its focus on manufacturing, physical infrastructure investments, and locations across Republican-voting states like Kentucky and Texas. Additionally, Apple CEO Tim Cook met with President Trump personally. Collectively, Apple understands that doing business in the US now involves a lot of marketing, public relations, and an emphasis on manufacturing and physical buildings rather than digital or intangible assets. 

Collectively, these two examples demonstrate that the US government is increasingly intervening in corporate decision making. Seen in this broader context, the US government’s seemingly overt intervention in some corporate boardrooms is therefore likely to continue as the tech race with China accelerates. For tech companies with prominent CEOs, particularly those with perceived connections to China, it would be prudent to assess how they might take the initiative to mitigate negative political attention from the US government.

Analysis

The US and Chinese governments jointly announced another 90-day extension of tariff relief measures, moving the deadline to November 12; both sides appear optimistic about finding common ground. This proverbial kicking of the can has also created a semi-stable trade environment, insomuch as companies can continue to anticipate the current tariff rates, which will enable businesses to make informed decisions. Given President Trump’s publicly stated desire to meet with Xi Jinping by the end of the year, however, the US has tipped its negotiating hand and any significant progress is also unlikely. Instead, Beijing will bet that the US will give it what it wants in order to secure the meeting of the two heads of state. In this scenario, US companies will suffer the potential consequences of the president’s desire to make a deal.

While China and the US continue to negotiate at the state-to-state level, US businesses also face an evolving regulatory landscape that has led to unusual developments. After the government decided to rescind its export controls on sales of Nvidia’s H20 chips to Chinese markets, it failed to issue any of the relevant licenses for over a month. In a deal that can only be described as greasing the wheels, Nvidia CEO Jensen Huang met with President Trump and promised the government a 15% cut of the company’s revenue made from the sale of H20 chips to China. While chip-producer AMD has yet to publicly comment, it has also acceded to a similar revenue-sharing agreement. It is perhaps a bitter irony, then, that China is urging its companies to avoid purchasing US chips, for fear they contain a backdoor that will steal their IP, and industry voices are questioning if the high-price for a “castrated product” is worth their money.

All signs seem to suggest that the US government is taking a page from its negotiating adversary when it comes to intervening in the operations of private businesses. The Nvidia deal, for instance, comes only weeks after the US government agreed to take a “golden share” in Nippon Steel as a condition of approving Nippon Steel’s acquisition of US Steel. If President Trump continues insisting on personally striking deals with major corporations as a precondition for permitting them to do business, we may be entering a new era of state capitalism with US characteristics. This is not to suggest that the US government is taking a socialist path–whereby the state owns the means of production–or becoming like other, well-known state-led markets, such as Russia or Brazil. It is instead meant to highlight the US government’s significant departure from its former faith in free-market capitalism and to alert businesses to the very real possibility that they may need to consider the government’s interests in their future expense projections.

Analysis

Huawei launched a new software tool called UCM (Unified Cache Manager) that manages AI memory and reportedly makes Huawei’s large language models (LLMs) run faster and more cheaply. The announcement marks another milestone in China's systematic campaign to develop alternatives to US-controlled technologies, demonstrating Beijing's commitment to what it calls "technological self-reliance."

Chinese business media reported that the new approach could lessen reliance on high-bandwidth memory (HBM) chips for inference, although we are still awaiting technical details and third-party benchmarks. The timing is particularly significant, given that HBM is critical for running large AI models and delivering the necessary high bandwidth and low latency. However, China currently faces comprehensive restrictions on HBM exports from the United States and its allies, effectively limiting access to newer versions that are essential for cutting-edge AI applications.

For US policymakers, Huawei's UCM announcement poses uncomfortable questions about export control effectiveness. The US Commerce Department issued guidance stating that the use of Huawei's Ascend artificial intelligence chips "anywhere in the world" violates the government's export controls. Yet rather than constraining Chinese AI development entirely, restrictions appear to be spurring domestic innovation efforts.

The broader implications extend beyond semiconductors to fundamental questions about technological leadership in the AI era. Huawei's focus on AI inference—the critical deployment phase where trained AI models apply their capabilities to real-world tasks—addresses a key performance bottleneck that affects the practical utility of LLMs. While it remains unclear whether UCM delivers measurable performance advantages over existing solutions, the company's ability to develop software-based alternatives to hardware constraints suggests a strategic shift from dependence to original innovation that could reshape competitive dynamics.

Chinese tech firms are leveraging software improvements to compensate for limited access to advanced hardware, as demonstrated by Huawei's UCM algorithm. However, achieving broad, system-level independence across chips, memory, EDA tools, and software ecosystems remains challenging. The current evidence supports steps toward reduced dependence rather than comprehensive technological autonomy.

Huawei’s announcement will likely be framed by Chinese officials as evidence of growing technological resilience. For US policymakers, it serves as a reminder that technological competition is evolving beyond simply restricting access to existing technologies—and whether restrictions are inadvertently providing motivation for competitor innovation to accelerate.

Analysis

As US export controls reduce the international supply amid skyrocketing demand for high-performance chips, semiconductor and chip manufacturing companies will almost certainly face an increasing barrage of attempted IP theft and chip smuggling attempts. Governments, especially China’s, see advancement in technology as critical to national security. Therefore, chip designers and manufacturers will continue to find themselves subject to geopolitical tensions over the contest for technological supremacy.  

Two recent cases demonstrate what is at stake for the chip industry. In one case, the FBI arrested two Chinese nationals for allegedly exporting tens of millions of sensitive microchips to China through their California-based company, ALX Solutions Inc., in violation of the US Export Control Reform Act. The case demonstrates the high risk facing US companies from other US-based entities that front for Chinese interests, and underscores the importance of adequate diligence on suppliers or business partners. 

In a second case, Taiwan Semiconductor Manufacturing Co. (TSMC), the world’s largest chip maker, fired and indicted several employees for attempting to steal sensitive information about the company’s 2-nanometer chip development and production. The TSMC case also highlights the multivariate nature of industrial espionage. Taiwanese media reported that the stolen IP was allegedly leaked to TSMC’s Japanese competitor, Rapidus, where the chips were used to help adjust a product in development. One of the accused individuals was also employed by a Japanese company, Tokyo Electron Ltd., which is a supplier to TSMC. In response, Tokyo Electron announced that it had dismissed the employee and was cooperating with authorities. However, the public relations damage to the company already appears to be a fait accompli.

Both examples highlight that industrial complacency about suppliers can be devastating. Just because a company is registered in the US or an ally like Japan, does not eliminate risk, even if the risk profile is lower than in countries like China. For the tech sector in particular, these cases also demonstrate that threats emanate from two major sources: Competitors and governments. The two categories differ significantly in that domestic and allied competitors are subject to the criminal and civil legal system, while governments like China, can evade the consequences of both. That said, competitors will continue to offer financial rewards for industrial secrets, and governments will pressure and induce tech employees to divulge corporate information. The two dynamics overlap where governments assist domestic companies to “poach” skilled talent. This threat applies beyond semiconductors to AI, biotech, robotics, energy, mining and other sectors. Companies that fail to prepare adequately against such attacks are therefore likely to face dire economic consequences.