In the proclamation, “Covered Products” are defined as “certain advanced computing chips and certain derivative products described in the Annex to this proclamation.” Publicly available text of the proclamation does not yet reproduce that Annex or its detailed list of tariff lines, but it indicates these are high‑end AI/data‑center chips and closely related products. As of now, the only official place the Annex is referenced is at the end of the White House proclamation page itself; typically, the detailed list of products will also be published either as a separate Annex document linked from that page or in a follow‑on notice in the Federal Register amending the Harmonized Tariff Schedule of the United States (HTSUS). In short: the exact line‑by‑line list of “Covered Products” is not yet accessible in the public text, and you would need to check the White House proclamation page and future Federal Register/HTSUS updates for the Annex once posted.
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 product are entering the United States “in such quantities or under such circumstances as to threaten to impair the national security.” If Commerce finds a threat, it sends a report to the President. The President then has broad legal authority to “adjust the imports” of that article and its derivatives, which can include imposing tariffs or other trade restrictions, adopting a plan of action with adjustments over time, and negotiating agreements with other countries to remove the threat.
The proclamation explicitly cites this authority, and CRS and other legal summaries confirm that Section 232 is the statutory basis for past steel and aluminum tariffs and similar “national security” duties.
The proclamation does not spell out a precise test for when imports “contribute to the buildout of the United States technology supply chain,” but it does say:
• The 25% tariff does not apply when Covered Products are imported for specified uses: U.S. data centers, repairs/replacements in the U.S., U.S.-based R&D involving these chips, U.S. startups, non–data‑center consumer uses in the U.S., non–data‑center civil industrial uses in the U.S., U.S. public‑sector applications, and “other uses that the Secretary determines” strengthen the U.S. tech supply chain or domestic semiconductor‑derivative manufacturing. • The Secretary of Commerce is explicitly charged with making those “other uses” determinations, “consider[ing] factors he deems relevant, including the need to address the national security threat … and the purpose of this proclamation.” • The Secretary, in consultation with the International Trade Commission Chair and the CBP Commissioner, is also tasked with deciding what “end‑use certifications or other administrative measures are necessary” and publishing any such rules in the Federal Register, which is how end‑use determinations will be operationalized at the border.
In practice, importers should expect a formal Commerce/CBP process (such as end‑use or end‑user certifications attached to entries, similar to past Section 232 regimes) to document that their shipments fall into one of the exempt uses; the specific criteria will only be clear once those regulations or guidance are issued.
An “ad valorem” duty is a tariff charged as a percentage of the customs value of the imported goods, rather than a fixed amount per unit or per weight. “Ad valorem” literally means “according to value.”
For this proclamation, a 25% ad valorem duty means CBP will calculate 25% of the declared customs value of each shipment of Covered Products (usually based on the transaction value adjusted under customs valuation rules) and collect that amount as duty when the goods are entered for consumption. If, for example, the dutiable customs value of a shipment is $1 million, the Section 232 ad valorem duty would be $250,000 in addition to any other applicable duties and fees.
The proclamation only sketches the concept of a tariff‑offset program and does not detail how it will work, but it follows the pattern of other recent Trump‑era “offset” schemes:
• The proclamation says that after negotiations, the President “may consider imposing significant tariffs” on a broader range of semiconductors and related products and that this broader tariff “be accompanied by a tariff offset program to enable companies investing in United States semiconductor production and certain parts of the United States semiconductor supply chain to obtain preferential tariff treatment.” • Under the 2025 auto‑parts Section 232 action, Commerce created an “auto tariff offset” process: automakers or suppliers that made or committed to specific U.S. investments could apply for credits or reduced effective tariff rates on their covered imports, based on certified investment plans, spending, or sourcing changes. Companies applied through a Commerce‑run portal, submitted documentation of qualifying investments, and then received preferential tariff treatment (often via special tariff classifications or refunds) on later entries that met the program’s conditions.
By analogy, a semiconductor tariff‑offset program would likely: • Define what counts as qualifying investment (e.g., building or expanding U.S. fabs, packaging facilities, R&D centers, or key supply‑chain nodes). • Require firms to apply to Commerce with detailed investment plans and timelines. • Grant preferential tariff treatment (lower or zero Section 232 rates, or credits) on imports that are tied to those approved investments, administered via specific HTSUS subheadings and CBP procedures. The exact design will only be known once Commerce issues implementing rules after the broader tariffs are decided.
The proclamation gives CBP primary responsibility to administer the new tariff, and—based on how past Section 232 tariffs have been enforced—importers should expect the following:
• HTSUS changes and special tariff numbers: The Secretary of Commerce, in consultation with the ITC Chair and the CBP Commissioner, must determine and publish any HTSUS modifications and “end‑use certifications or other administrative measures” needed to implement the tariff. In practice, this usually means creating new Chapter 99 tariff lines for the Covered Products, with instructions in the Federal Register. • Entry procedures: For each shipment, importers will have to declare the appropriate HTSUS code (including any new Section 232 provision), report the customs value, and pay the additional 25% duty on entries that do not qualify for an exemption. • End‑use certifications for exemptions: Because the duty does not apply when the chips are imported for specific uses (data centers, R&D, repairs, startups, specified consumer/industrial and public‑sector uses, or other qualifying uses), CBP and Commerce are likely to require end‑use certifications or statements at entry or post‑entry, similar to the documentation CBP uses in the steel and aluminum Section 232 programs. • Foreign‑trade zones (FTZs): The proclamation requires that Covered Products admitted into FTZs on or after the effective date (unless in domestic status) must be in “privileged foreign status,” meaning the Section 232 duty rate is locked in and applies when the goods enter U.S. commerce. • No drawback: The proclamation explicitly bars duty drawback on these Section 232 duties, so importers cannot reclaim them when re‑exporting.
CBP will operationalize these points through Cargo Systems Messaging Service (CSMS) notices and updated guidance once the HTSUS changes and certification requirements are finalized.
Economic impacts and timing are still uncertain, but existing analysis of semiconductor tariffs and the structure of this proclamation point to several likely effects:
• U.S. data‑center and AI firms: Because most data‑center and R&D uses are exempt, the immediate direct cost hit to U.S. hyperscalers and AI labs may be limited, provided they can document qualifying end uses. However, any chips or derivative products that do not fit the exempt categories will face higher landed costs, potentially slowing projects that rely on non‑exempt configurations or intermediates. • U.S. manufacturers and device makers: For firms using Covered Products in non‑exempt industrial or commercial equipment, the 25% duty will raise input costs unless they qualify for exemptions or future tariff‑offset credits. Past modeling of hypothetical broad semiconductor tariffs suggests higher production costs could reduce U.S. output and competitiveness in electronics, autos, and other chip‑intensive sectors, and could encourage more production outside the U.S. where these tariffs do not apply. • Consumers: To the extent that costlier imported chips flow into finished goods sold in the U.S. and are not exempt, prices of some electronics, vehicles, and other digital products could rise. Studies of earlier chip tariffs and the U.S.–China trade war find that much of the tariff burden tends to be passed through to U.S. buyers in higher prices or reduced product choice. • Global supply chains: Tariffs targeted at advanced AI chips could accelerate efforts by non‑U.S. firms and some U.S. buyers to shift supply chains or design products to avoid the specific HTSUS lines covered by the tariff, potentially fragmenting global chip trade further. Analyses from industry and economic research groups generally find that broad chip tariffs would be net‑negative for U.S. growth, even if they encourage some additional U.S. fabrication investment. • Timing of broader tariffs: The proclamation directs Commerce and USTR to report back within 90 days on negotiations and says the President “may consider imposing significant tariffs” on a wider range of semiconductors and related products after that. It does not commit to a specific date or guarantee that broader tariffs will occur; their scope and timing will depend on the outcome of those negotiations and on Commerce’s ongoing monitoring.
Overall, research on semiconductor tariff scenarios indicates that while targeted measures can pressure foreign suppliers and may modestly support U.S. fab investment, they also risk higher costs and slower innovation across many downstream U.S. industries that rely on complex, globalized chip supply chains.