Chip Tariff War: U.S. Precision Strikes and the Cracks in the Global Supply Chain

On January 15, 2026, the morning in Washington was disrupted by a presidential proclamation. Citing Section 232 of the Trade Expansion Act, the White House formally imposed a 25% import tariff on specific high-performance semiconductor chips. This measure, the result of a nine-month investigation, directly targets core components driving the artificial intelligence revolution, such as NVIDIA's H200 and AMD's MI325X. The proclamation text calmly states a fact: the United States currently can only fully and independently produce approximately 10% of the chips it requires. This deep reliance on foreign supply chains constitutes a significant economic and national security risk.

However, this is not an indiscriminate trade strike. A careful reading of the terms reveals that chip imports for U.S. data centers, startups, non-data center civilian industrial applications, and the public sector are cleverly exempted. The sharp edge of the tariffs falls precisely on those high-end chips that transit through the United States and ultimately flow overseas, especially to China. In his subsequent remarks, President Trump described this mechanism more bluntly: We allow them to export, but the United States must receive 25% of the sales. It's a good deal.

This game surrounding silicon wafers extends far beyond the mere increase or decrease of tariff numbers. It is a recalibration of the global technological power structure, a strategic adjustment made in the name of national security but aimed at reshoring industries, and it may very well serve as a milestone in the process of technological decoupling between the United States and China.

"Precision Surgery" under the terms

Analysis reveals that the design and implementation of this tariff action exhibit a high degree of complexity and strategic planning, markedly different from previous crude trade protection measures.

First and foremost, the choice of legal basis is quite thought-provoking. The U.S. government abandoned the commonly used Section 301, which is typically employed to address unfair trade practices, and instead invoked Section 232, which primarily targets national security threats. This shift elevates the issue from one of fair competition to a matter of existential threat, granting the administration greater discretionary power and making it more challenging for trade partners to contest its legitimacy. In a briefing, White House Chief of Staff Will Scharf explicitly linked this move to the goal of ensuring the security of the U.S. international supply chain, with the national security narrative serving as the underlying theme for all actions.

The scope of the tariffs resembles a meticulously calibrated surgical procedure. It does not cover all semiconductors; instead, through an annex list, it is strictly limited to cutting-edge products meeting specific high-performance standards (such as computing power thresholds). Chips like NVIDIA's H200 and AMD's MI325X, designed specifically for large-scale AI training and inference, are the first to be affected. More critically, the tariff trigger conditions are tied to the chip's journey: only those manufactured overseas (primarily at TSMC's wafer fabs in Taiwan) and then transshipped through the United States to a third country (particularly China) are subject to the heavy tariffs. Similar products imported directly into the United States for domestic consumption are exempt.

This mechanism of taxing upon transit creates an unprecedented model of regulation and revenue generation. According to the new regulations, all specific high-end chips destined for sale in China must be diverted from their manufacturing origin in Taiwan to the United States, where they undergo inspection by third-party laboratories. Once they enter U.S. customs territory, a 25% tariff automatically takes effect. Donald Trump described it as the United States taking a 25% cut from the sales revenue. This arrangement has sparked sharp legal challenges: The U.S. Constitution explicitly prohibits taxes on exports, so does this fee, levied on goods in transit and ultimately borne by foreign buyers, essentially constitute an unconstitutional export tax? Despite the controversy, this mechanism has rapidly evolved from concept to practice, opening up a new trickle of revenue for the U.S. Treasury from technology trade.

The complex exemption system reveals the inherent contradictions and pragmatic considerations within the policy. On one hand, the U.S. government aims to reduce its reliance on foreign chip manufacturing; on the other hand, it soberly recognizes that completely cutting off the supply of high-end chips would immediately paralyze its own booming AI industry and innovation ecosystem. Therefore, creating exemption channels for key domestic users such as data centers, R&D institutions, and the public sector has become a safety valve for maintaining its own technological competitiveness. Commerce Secretary Howard Letnik has been granted broad authority to adjudicate exemptions, further increasing the flexibility and uncertainty in policy implementation. Zhang Zhiyang, a tax expert at KPMG, points out that the specific criteria for determining exempt uses are not yet fully clear, making it difficult for companies to accurately assess their own situation at present, highlighting the ambiguities in the policy during its initial implementation phase.

Killing Three Birds with One Stone: A Multidimensional Interpretation of Strategic Intent

On the surface, this appears to be a trade policy; upon deeper analysis, however, it reveals multiple strategic objectives encompassing industry, geopolitics, and fiscal policy.

The most immediate goal is the reshoring of manufacturing and the restructuring of supply chains. The White House announcement made no attempt to conceal its intent: by increasing import costs, it aims to create incentives for chip manufacturers to shift more production capacity to the United States. The U.S. maintains a lead in semiconductor design (home to giants like NVIDIA, AMD, and Intel), but in the manufacturing segment, particularly for cutting-edge processes, it heavily relies on Asian foundries such as TSMC. This geographical separation of design and manufacturing is seen as a core vulnerability. The tariff policy, coupled with the massive subsidies from the CHIPS and Science Act, forms a carrot-and-stick combination aimed at reversing this situation. Trump claimed in his speech that since his return to the White House, the U.S. has secured $18 trillion in investment commitments. Although this figure requires verification, it indeed reflects his strategy of using tariffs as leverage to attract capital backflow.

The leverage of geopolitics is particularly prominent. The policy is clearly targeted at China, yet it differs from a comprehensive embargo. It allows companies like NVIDIA to sell H200 chips to vetted Chinese clients, provided they pay a high "toll fee." This essentially creates a controlled dependency: it prevents China from easily obtaining the most advanced computing power to accelerate its AI militarization applications, avoids completely cutting off supply which might force China into a desperate, all-out push for self-reliance, while simultaneously profiting from China's demand. The Trump administration appears to be attempting to find a delicate balance between strangling and doing business. Furthermore, the policy sends a clear signal to chip-producing countries and regions such as Taiwan, South Korea, and Japan. The announcement implies that the U.S. will proactively initiate negotiations with these partners. If agreements ensuring supply chain support for U.S. domestic production capacity cannot be reached within 180 days, the scope of tariffs may be further expanded. This undoubtedly increases the U.S.'s bargaining chips in trade and security negotiations with its allies.

Generating Fiscal Revenue and Implementing Transactional Trade Policies. Trump views tariffs as a significant source of revenue, repeatedly claiming they have brought hundreds of billions of dollars to the United States. Imposing tariffs on transiting high-end chips represents the ultimate application of this concept in the technology sector. It directly transforms the global urgent demand for AI computing power into U.S. Treasury revenue. This approach, which highly fiscalizes and transactionalizes trade policy (we get 25%), overturns the traditional rules-based multilateral trade philosophy, leaning more towards a power-based commercial profit-sharing model.

Paving the way for broader actions in the future. The White House fact sheet clearly states that this is only the first step in a two‑step process. The current tariff scope is relatively focused, but it may expand to cover a wider range of semiconductors and their derivatives in the future. This leaves room for subsequent policy escalation and keeps the global semiconductor industry in a state of ongoing uncertainty. This sword‑of‑Damocles strategy itself exerts a powerful deterrent and guiding effect on corporate investment and supply‑chain planning.

Chain Reaction: The Shocks and Games of the Global Industrial Chain

A stone thrown into a pond sends ripples far and wide. The chain reaction triggered by this U.S. move is rapidly spreading across the global map of technology and trade.

The most direct bearers are multinational chip giants. The attitudes of NVIDIA and AMD are quite telling. An NVIDIA spokesperson publicly praised Trump's decision, calling it a great balancing act by the United States to support high-wage jobs and manufacturing. This is clearly a pragmatic statement made under established policy. Obtaining a license to sell the H200 to China, despite incurring additional costs, is better than completely losing this vast market. AMD cautiously stated it complies with all U.S. export control laws and policies. The slight dip in their stock prices in after-hours trading reflects the market's weighing of short-term cost increases against long-term market access. The real challenge for these companies lies in how to reconfigure their extremely complex global logistics chains to adapt to the new Taiwan-U.S. (inspection, taxation)-China route, while also preparing for the risk of future policy potentially tightening further.

Taiwan and TSMC are placed under the geopolitical spotlight. As the absolute main force in global high-end chip foundry, TSMC is the implicit key target of this tariff policy. The United States hopes to reduce its reliance on TSMC's manufacturing but cannot completely break away in the short term. This contradiction makes TSMC's position exceptionally delicate. It must carefully balance the pressure from the U.S. to shift production (such as accelerating the construction of the Arizona factory) with its own considerations regarding global layout and commercial interests. Mainland China has accused the United States of economic plunder by exerting pressure on TSMC, further deeply entangling the already highly sensitive chip manufacturing industry with the situation in the Taiwan Strait.

China's response strategy will be a key variable affecting the overall situation. Faced with the U.S. toll strategy, China cannot remain indifferent. Reports indicate that the Chinese government is drafting rules to control the quantity of overseas semiconductors (especially high-end AI chips) purchased by domestic enterprises. This reflects Beijing's dilemma: on one hand, it does not want domestic AI development to fall behind due to a shortage of computing power; on the other hand, it hopes to use this opportunity to support local chip industries such as SMIC and HiSilicon, reducing external dependence. China's countermeasures may not be limited to reciprocal tariffs (as China imports few similar high-end chips from the U.S.), and are more likely to be reflected in export controls on critical raw materials like rare earths, market access restrictions for U.S. technology companies, or adopting a more assertive stance on other geopolitical issues. A tug-of-war of precise decoupling and counter-decoupling around cutting-edge technology has already begun.

The global semiconductor industry landscape is facing restructuring pressures. U.S. policies are forcibly altering the economic geography of the semiconductor supply chain. The previously efficiency-optimized model of globalized division of labor is giving way to a trend of regionalization driven by national security and regional self-sufficiency. Regions and countries such as Europe, Japan, and South Korea are all increasing investments in their domestic chip industries. This tariff initiative launched by the United States is likely to accelerate the formation of multiple relatively independent yet mutually competitive semiconductor industry ecosystems worldwide. For global technology companies, the stability and cost of supply chains will face long-term challenges, forcing them to pay a higher premium for security and prepare multiple sets of contingency plans.

The Crossroads of the Future

This chip tariff action in early 2026 is by no means an isolated event. It represents a deep implementation of the Trump administration's "America First" trade philosophy in the most strategically significant technology sector. It also marks a new phase in the U.S.-China strategic competition, evolving from the trade war and technology war toward a more fundamental battle over computing power.

Whether the policy will achieve its intended effects remains highly uncertain. Can a % tariff truly drive large-scale reshoring of cutting-edge chip manufacturing to the United States? Given the hundreds of billions of dollars in investment, lengthy timelines, and shortage of specialized talent required to build advanced wafer fabs, the answer is far from straightforward. It may prompt some capacity relocation, but it is unlikely to reshape the global industrial landscape in the short term. Instead, it is more likely to increase the cost of global research and development, ultimately borne by consumers and tech companies worldwide.

Challenges at the legal level loom like the Sword of Damocles. The U.S. Supreme Court is currently hearing lawsuits challenging the president's tariff authority, with Trump claiming that those filing the suits are individuals who are very pro-China. If the Supreme Court ultimately rules that such applications of Section 232 are unconstitutional, the entire policy framework could be shaken instantly. Even if legally sound, this trade policy, which heavily relies on executive orders and is filled with discretionary power, injects immense unpredictability into the international business environment.

From a broader perspective, as chips—the crude oil of the information age—become the core focus of competition among nations, the world is sliding toward a more fragmented technological system. The global flow of knowledge and collaborative networks that innovation relies on may be severed by tariff walls and regulatory barriers erected in the name of security. Through this precise surgical strike, the United States aims to secure its leading position in AI, profit from the technological demands of its rivals, and reshape the global supply chain. This is a risky move, and an ambitious one.

The ultimate outcome will not only determine the commercial fate of giants like NVIDIA and TSMC but also shape the trajectory of global artificial intelligence development and the balance of power in geotechnology over the next decade. Every participant in the global industrial chain must now reassess their position and prepare for a new era characterized by higher costs, greater risks, and increased fragmentation. The number of transistors on chips continues to grow according to Moore's Law, yet the global system that carries these chips may be heading in an unknown direction, contrary to the trends of globalization.