White House invokes Section 232 and imposes 25% tariff on certain advanced computing chips to address national security risks

Section 232 of the Trade Expansion Act of 1962 (19 U.S.C. §1862) lets the U.S. Secretary of Commerce investigate whether imports of a particular “article” are being brought into the country in such quantities or under such circumstances that they threaten to impair national security. If Commerce makes an affirmative finding, the President then has 90 days to decide whether to concur and, if so, may “adjust imports” using broad tools such as tariffs, quotas, or other restrictions, and may negotiate agreements with foreign governments. These actions are justified on national‑security grounds, including both direct defense needs and the broader health of the U.S. economy and critical industries.

The proclamation imposes an immediate 25% ad valorem duty on “certain advanced computing chips and certain derivative products” defined as “Covered Products” in an Annex to the proclamation. Publicly available documents (the fact sheet and proclamation text) do not list the full model‑by‑model or tariff‑line coverage of that Annex, so the precise set of chips and derivative products is not yet known from open sources.

The White House fact sheet and related reporting instead describe the measure as a “very narrow category” of advanced AI/data‑center semiconductors and cite Nvidia’s H200 AI processor and AMD’s MI325X accelerator as illustrative examples of the kinds of high‑end chips affected, not as the only models covered. These particular chips are referenced because they are among the flagship, highest‑performance AI accelerators currently used in large data centers and thus central to the U.S. AI and semiconductor supply‑chain policy the proclamation targets.

The administration defines the exemption in terms of end use rather than specific chip models. Under the proclamation, the 25% duty on “Covered Products” does not apply when the imported chips and covered derivatives are used for:

• U.S. data centers;
• Repairs or replacements performed in the United States;
• Research and development in the United States;
• U.S. startups;
• Non‑data‑center consumer applications in the United States;
• Non‑data‑center civil industrial applications in the United States;
• U.S. public‑sector applications; or
• Other uses that the Commerce Secretary determines contribute to “the strengthening of the United States technology supply chain or domestic manufacturing capacity for derivatives of semiconductors.”

To apply this, Commerce (in coordination with the U.S. International Trade Commission and U.S. Customs and Border Protection) is directed to set up any needed Harmonized Tariff Schedule changes, end‑use certifications, or other administrative measures, likely meaning importers will need to certify qualifying end uses to claim the exemption.

A “tariff offset program” is a mechanism that lets firms reduce or avoid some of the tariffs they would otherwise owe, if they meet specified conditions such as investing in U.S. production. In the semiconductor proclamation, Commerce recommended that any future, broader Section 232 tariffs on semiconductors be paired with a tariff‑offset program so that:

• Companies investing in U.S. semiconductor fabrication or key segments of the U.S. chip supply chain could obtain preferential tariff treatment (e.g., credits, rebates, or reduced rates) on their covered imports.

The proclamation does not spell out the detailed structure, but it explicitly links the concept to an earlier auto‑sector Section 232 tariff‑offset process, where U.S. automakers could apply for offsets to auto‑parts tariffs based on the volume and value of their U.S. production. By analogy, a semiconductor tariff‑offset program would likely condition relief on demonstrable domestic investment, capacity, or employment in semiconductor manufacturing and related supply‑chain activities.

Timeline and enforcement for the initial 25% tariff and possible broader measures are laid out as follows:

• Effective date and duration of the 25% tariff: The proclamation states that imports of the covered advanced computing chips and derivatives “will be subject to a 25 percent ad valorem duty rate” starting with goods entered for consumption, or withdrawn from warehouse for consumption, on or after 12:01 a.m. Eastern Standard Time on January 15, 2026, and that the tariff will remain in effect unless it is later reduced, modified, or terminated.

• Near‑term negotiations and broader tariffs: Commerce and the U.S. Trade Representative are directed to pursue or continue negotiations with foreign jurisdictions and to report back to the President within 90 days. Depending on how those talks proceed, the President “may consider imposing significant tariffs” on a wider range of semiconductors, semiconductor manufacturing equipment, and derivative products, together with a tariff‑offset program.

• Ongoing review and possible adjustment: Commerce must monitor imports and provide an update on the market for semiconductors used in U.S. data centers by July 1, 2026, so the President can decide whether to modify the tariff. Commerce is also instructed to advise the President if circumstances indicate that further action or removal of the tariff is warranted.

• Enforcement: U.S. Customs and Border Protection is explicitly authorized to take “any necessary or appropriate measure” to administer the tariff, including applying the new rate at entry, using end‑use certifications, and treating covered goods in foreign‑trade zones as “privileged foreign status” so the tariff applies when they enter U.S. commerce.

Expected impacts, based on the proclamation and early analysis, are:

• U.S. companies (designers, cloud providers, large industrial users):

• For uses that qualify for exemptions (e.g., U.S. data centers, R&D, startups, consumer devices), U.S. buyers of high‑end AI chips should be largely shielded from the 25% tariff in the near term, limiting direct cost increases.
• Firms that import covered chips for non‑exempt purposes face higher costs, which may encourage them to shift more assembly, integration, or manufacturing activity into the United States to meet exemption criteria or benefit from any future tariff‑offset program.

• Foreign suppliers and overseas fabs:

• Major overseas foundries and packaging operations (e.g., in Taiwan and other Asian economies) producing advanced AI accelerators for the U.S. market face a tariff barrier for non‑exempt U.S. uses, potentially reducing their price competitiveness versus chips made or packaged in the United States or by companies investing in U.S. capacity.
• Over time, this could re‑route some investment and production toward U.S. fabs and advanced‑packaging facilities if offset incentives are strong and predictable.

• Semiconductor supply chains:

• The action is intentionally narrow at first, targeting a small subset of advanced AI/data‑center chips, which should limit immediate disruption.
• However, the proclamation explicitly contemplates “significant” broader tariffs after negotiations conclude. If those are imposed on a wider range of chips or tools, they could materially raise import costs, complicate global sourcing strategies, and accelerate efforts by firms to diversify or localize supply chains in the United States and allied countries.

• Consumer prices for electronics:

• Because the initial tariff is focused on high‑end AI accelerators, not mass‑market CPUs, smartphones, or PCs, consumer electronics prices are unlikely to see large immediate, direct increases.
• If the policy later expands to broader categories of semiconductors or derivative products used in consumer devices, higher component costs could gradually pass through into more expensive electronics, though the magnitude would depend on the final tariff scope and on how easily firms can reorganize supply chains.

Commerce’s Section 232 investigation into semiconductors, semiconductor manufacturing equipment, and their derivative products reached several key findings that underpin the “national security” rationale:

• Insufficient U.S. production and heavy import dependence: The United States consumes about one‑quarter of the world’s semiconductors but “currently fully manufactures only approximately 10 percent” of the chips it requires, leaving it “heavily reliant on foreign supply chains” and creating a “significant economic and national security risk.”

• Critical to defense and military systems: Semiconductors are “essential to the United States’ economic, industrial, and military strength.” Modern defense systems rely on high‑performance chips for radar, communications, electronic warfare, cybersecurity, and guidance and control of missiles and drones, including specialized chips that can withstand extreme environments.

• Foundational to critical infrastructure: Semiconductors are described as essential to all 16 U.S. critical‑infrastructure sectors (communications, energy, nuclear, healthcare/medical, etc.), underpinning functions like smart‑grid controls, nuclear‑reactor safety systems, communications networks, and medical devices.

• Capacity too low for defense and commercial needs: U.S. manufacturing capacity is “too low to meet projected national defense needs” and the requirements of a growing commercial industry. A disruption of import‑reliant supply chains could “strain the United States’ industrial and military capabilities.”

• AI‑enabling chips as a specific concern: AI‑enabling semiconductors are identified as an “important element of many data centers,” and the investigation concludes that imports of these AI chips, when not used to build out the U.S. technology supply chain, in current quantities and circumstances pose a threat to national security.

These findings led Commerce to recommend a two‑phase response, starting with the narrow 25% tariff on selected advanced computing chips and potentially moving to broader tariffs plus a tariff‑offset program.