The U.S. said that chip tariffs effectively encourage shifting chip production to the country.

The United States continues to use tariffs as a tool to support local semiconductor production

U.S. Trade Representative Jamieson Greer recently stated that federal authorities view import duties on chip components as an effective way to stimulate domestic manufacturing. At the same time, the government does not plan to change tariffs significantly in the near future.

Key points

When and how Greer explained Current government position at the Micron Technology plant expansion event in Northern Virginia “We continue dialogue with industry about the size and timing of new tariffs, but nothing is introduced tomorrow or next week” Tariff goal Simplify bringing production back to the U.S. “It needs to be done on time and in the right volume”

Context

- In January the Department of Commerce confirmed that dependence on imported semiconductors poses a national security threat.
- President Donald Trump instructed negotiations with key exporters about tariffs.
- After Trump’s recent meeting with the Chinese president, the U.S. position remained unchanged, as Greer noted.

How tariffs work

Preferential rates are applied only to companies that will localize production within the country in the future. This allows importing some components at reduced duties on the condition that they are subsequently manufactured in the United States.

Micron and future plans

- The company plans to invest $200 billion in memory development in the U.S. over the next 20 years.
- With growing demand for AI memory, new capacity is coming online slower than the need.
- Competitors Samsung and SK Hynix may also face tariffs if they do not expand production in the U.S.

Micron’s head explained that building a new plant takes several years; the speed of outfitting depends on demand forecasts. As CEO Sanjay Mehrotra noted, the key factor is readiness to meet market demand. The company is currently securing long‑term contracts with customers to increase memory supply predictability.