In a sweeping policy reversal that has sent ripples through global markets, the Trump administration in January 2026 shifted the United States’ approach to advanced semiconductor exports, allowing the sale of cutting-edge AI chips to China under stringent new conditions. The move, which replaces a two-year-old blanket ban with a more nuanced, transaction-by-transaction review, comes amid mounting economic and strategic tensions between the world’s two largest economies—and has already triggered a cascade of consequential reactions in both Washington and Beijing.
According to InvestingLive, the US Department of Commerce’s Bureau of Industry and Security formally changed its export license review policy on January 13, 2026, moving from a “presumption of denial” to a “case-by-case review” for advanced chips like NVIDIA’s H200 and AMD’s MI325X equivalents. Just a day later, the Trump administration announced it would approve sales of NVIDIA’s H200 chips to China, albeit with a hefty 25% tariff and a 50% volume cap in place. The administration insisted that these changes would allow the US to monetize Chinese demand while maintaining a tight grip on technology and national security, but critics in Congress and beyond have sounded alarms about the implications for America’s competitive edge in artificial intelligence.
Chinese tech giants wasted no time responding. As InvestingLive reports, firms such as Alibaba, Tencent, and ByteDance placed orders for more than two million H200 chips, representing a potential windfall of up to $14 billion for US semiconductor companies in 2026 alone. ByteDance, the parent company of TikTok, has made H200 acquisition a core pillar of its AI research strategy for the coming year. Yet, despite this apparent opening, the chips remain subject to rigorous transaction-specific risk assessments and a mandatory third-party US testing phase before export. Transactions linked to China’s military, intelligence, or surveillance sectors are still subject to strict controls.
The new US policy comes at a time of unprecedented mutual dependence and strategic gamesmanship between Washington and Beijing. Since December 2024, China has imposed outright export bans on critical minerals essential for semiconductor manufacturing—including silver and rare earth elements like gallium, which China almost entirely dominates in global refining. According to InvestingLive, China now refines about 70% of the world’s silver used in chips, giving it enormous leverage over the US and its allies. These export controls have put pressure on the US to soften its stance on chip exports, as American manufacturers face potential bottlenecks in their own supply chains.
Yet, the risks of the new policy are substantial. As InvestingLive explains, the H200 is not NVIDIA’s most advanced chip, but it is six times more powerful than any US chip currently available in China or produced by domestic Chinese firms such as Huawei. With Chinese AI developers now able to access millions of these chips, the gap in computing power between the US and China could narrow dramatically. If the US had continued to block all advanced chip exports, its compute capacity would have remained ten times greater than China’s in 2026—a lead that now appears at risk.
Congress has not taken these developments lightly. On January 22, 2026, the House Foreign Affairs Committee passed the AI Overwatch Act, which seeks to expand congressional oversight over AI chip exports. If enacted, the bill would require both the House committee and the Senate Banking Committee to approve any advanced chip shipments within 30 days, and could revoke existing export licenses at any time. This introduces a new layer of uncertainty for the global supply chain, as licenses granted by the Commerce Department could be overturned by lawmakers at a moment’s notice. The bill is expected to pass the House but may face resistance in the Senate, according to forecasts cited by InvestingLive.
Meanwhile, Beijing is playing its cards carefully. In early February 2026, Chinese regulators advised the country’s major financial institutions to pare down their holdings of US Treasuries, citing “concentration risks” and “market volatility.” As reported by InvestingLive, the directive was communicated to some of China’s biggest banks just before a phone call between President Trump and Chinese leader Xi Jinping, in which Trump outlined plans to visit China in April. The order does not touch China’s official state holdings of Treasuries, but it marks a continuation of a long-term trend: China’s direct Treasury holdings have fallen to $682 billion as of January 2026, a 17-year low, down from a peak of $1.3 trillion a decade ago. The report notes that China has become more of a “stealth seller,” quietly reducing its exposure rather than dumping assets outright.
The market has already felt the tremors. On February 9, 2026, 10-year US Treasury yields nudged up to 4.24%, from 4.22% earlier in the day, in response to news of China’s reduced appetite for US government debt. The underlying rationale, according to Beijing officials cited by InvestingLive, is that US assets have lost their appeal as a safe haven amid ongoing market turbulence and what they describe as the Trump administration’s “incoherent and erratic policy approach.”
Notably, despite the US policy shift, China has instructed its customs authorities to block imports of H200 chips and has warned national technology companies against purchasing them unless absolutely necessary. This move, reported by InvestingLive, suggests that Beijing is wary of becoming strategically dependent on US technology—especially technology that could be weaponized in a future conflict. China appears to be weighing the benefits of acquiring advanced computing power against the risks of deepening its reliance on American suppliers.
The international ramifications are equally complex. The Trump administration’s reversal risks undermining multilateral export control regimes negotiated with allies in the Netherlands, Japan, and South Korea. If these partners perceive US commitments as negotiable under industry pressure, they may reconsider their own restrictions, potentially eroding the united front that has characterized Western technology policy toward China since 2022.
Looking ahead, the implications are profound. In the short term, the decision is likely to accelerate China’s ability to build sophisticated AI models and scale up its data centers, narrowing the technological gap with the US. American companies like NVIDIA stand to benefit commercially, at least in the near future. In the longer term, however, experts warn that China could leverage H200 computing power for military applications—ranging from AI-enabled drones and precision strike systems to enhanced cyber warfare capabilities. Chinese cloud providers, armed with these chips, may soon be able to compete head-to-head with US giants such as Google Cloud and AWS.
With the global supply chain in flux and both sides holding strategic chokepoints, this new era of US-China technology relations is shaping up to be less about unilateral restrictions and more about transactional bargaining. As each nation seeks to protect its interests while exploiting the other’s vulnerabilities, the world is left to watch—sometimes with bated breath—as the future of AI, finance, and security hangs in the balance.