Trump imposes 25% tariff on advanced AI chips, citing national security

The White House announced that President Donald Trump imposed a 25% tariff on “certain advanced computing chips,” citing national security in a presidential proclamation and an accompanying fact sheet issued on Jan. 14, 2026. The administration named examples of affected products, including Nvidia’s H200 AI processor and AMD’s MI325X, and invoked Section 232 of the Trade Expansion Act of 1962 as the legal basis for the measure.

The proclamation described the action as narrowly focused but warned that broader tariffs on semiconductors and derivative products could follow. The fact sheet set out a list of explicit carve-outs that will exclude some imports from the tariff: chips destined for U.S. data centers, startups, non-data-center consumer applications, non-data-center civil industrial uses, U.S. public-sector applications, and imports brought in to “support the buildout of the U.S. technology supply chain and the strengthening of domestic manufacturing capacity for derivatives of semiconductors.” The administration emphasized the supply-chain and manufacturing-support exemptions, but the fact sheet did not publish a detailed product list or the criteria companies will need to meet to qualify for those carve-outs.

Officials did not provide harmonized tariff schedule codes or an itemized product list, leaving manufacturers, cloud providers and importers uncertain about which shipments will be assessed. The operational rules for the carve-outs were also not immediately available, and the White House said further guidance may be forthcoming.

The administration framed the tariff as part of a broader strategy to incentivize semiconductor manufacturing in the United States and reduce dependence on foreign production. Many major chip designers are U.S.-based, but most advanced chips are produced abroad, frequently by contract manufacturers such as Taiwan Semiconductor Manufacturing Company. The White House said the levy is intended to strengthen U.S. technology supply chains and encourage onshore production of high-end chips.

The move also changes the commercial calculus for sales to foreign customers. The administration indicated that it will permit certain commercial exports of affected chips while collecting the 25% tariff on those shipments. Nvidia, in an emailed statement, praised the decision as allowing “America’s chip industry to compete to support high paying jobs and manufacturing in America.” The White House warned that additional tariffs could be applied more broadly in the near term.

The tariff carries immediate implications for industries reliant on advanced AI accelerators and for public-interest sectors that are beginning to deploy AI tools. Hospitals, medical research centers and health-tech startups that use cutting-edge processors for imaging, diagnostics and drug discovery could face higher hardware costs or delays if supply routes are disrupted or if exemptions prove hard to obtain. Cloud providers and data centers, which the fact sheet singled out for exemption, will be watching for formal guidance to avoid interruptions to services that health systems and universities depend on.

Beyond industry effects, the policy raises equity concerns. Smaller labs and community health organizations with limited procurement leverage may struggle to absorb price increases, potentially widening gaps in access to AI-driven care. The administration said the carve-outs include startups and imports that support domestic manufacturing, but without clearer rules the practical reach of those protections remains uncertain.

Trade officials signaled that more detailed Commerce Department guidance and a formal determination under Section 232 will follow. Companies, investors and health-sector stakeholders will need that guidance to assess cost, supply and deployment risks tied to the new tariff.