The United States is considering another round of tariffs on semiconductor imports, according to a report based on people familiar with discussions inside the administration. The proposal could extend beyond individual chips to products that contain semiconductors, including laptops, gaming consoles and data center servers. The plan remains under development and could change before any final decision is made.

The reported proposal comes after the administration imposed a 25 percent tariff on certain advanced computing chips in January. Those duties were introduced under national security powers and were accompanied by broad exemptions for products used in US data centers, research, startups, consumer applications and other activities considered supportive of the domestic technology supply chain. The January action also left open the possibility of broader semiconductor tariffs after negotiations with trading partners.

The new proposal therefore appears to be part of a wider attempt to encourage more semiconductor production inside the United States. The reported approach would potentially connect tariff relief for foreign manufacturers to commitments to invest in American chip production. That would make the policy less about taxing imports alone and more about using access to the US market as an incentive for companies to build manufacturing capacity domestically.

The central difficulty is that semiconductors are not an isolated product. They are embedded in computers, servers, telecommunications equipment, vehicles, game consoles and increasingly sophisticated artificial intelligence infrastructure. Expanding tariffs from chips to finished technology products could therefore affect a much larger part of the technology supply chain.

Tariffs Are Being Used to Encourage Domestic Chip Investment

The administration’s semiconductor policy is based on a concern that the United States remains heavily dependent on overseas manufacturing for important parts of the chip supply chain. American companies dominate areas such as chip design and advanced semiconductor technology, but much of the physical manufacturing takes place outside the United States. Taiwan, South Korea and other Asian economies remain important production centers, while semiconductor manufacturing also depends on international networks for equipment, materials and packaging.

The January semiconductor proclamation explicitly identified imported chips, semiconductor manufacturing equipment and related products as a national security concern. It established an initial 25 percent tariff on certain advanced computing chips while directing US officials to negotiate with foreign governments and consider broader measures. The administration also stated that future tariffs could be accompanied by a program giving preferential treatment to companies investing in US semiconductor production.

The latest reported proposal appears to build on that framework. Under the approach being considered, foreign companies could potentially reduce their exposure to tariffs by committing more capital to US manufacturing. Such a system would give multinational companies an incentive to place factories and other parts of the supply chain inside the United States rather than simply paying higher duties on imported products.

That could support the administration’s industrial policy objective, but tariffs alone cannot quickly create semiconductor capacity. New fabrication plants require large amounts of capital, specialised equipment, highly trained workers, reliable electricity and water supplies, and years of construction and qualification. The policy can influence investment decisions, but it cannot immediately replace existing Asian manufacturing networks.

The United States is already attempting to increase domestic semiconductor capacity through financial incentives and industrial policy. The combination of subsidies, tax incentives and tariffs suggests that Washington is trying to use several tools simultaneously rather than relying on import duties alone.

Extending Tariffs to Devices Could Increase the Economic Cost

The reported expansion from chips to finished technology products would be particularly important because it could widen the number of companies affected. A tariff applied directly to a semiconductor can raise the cost of a specific component. A tariff applied to a laptop, server or gaming console can affect the final product sold to consumers and businesses.

That distinction becomes especially important for data center servers. Artificial intelligence companies and cloud providers are investing heavily in computing infrastructure, increasing demand for advanced processors, memory and networking equipment. A broader tariff regime could therefore arrive while technology companies are attempting to expand computing capacity at an unusually rapid pace.

The January tariff structure illustrates why exemptions matter. The administration specifically excluded certain advanced chips used in US data centers and several other applications from the 25 percent duty. Those exemptions were designed to avoid placing an immediate burden on parts of the domestic technology buildout.

If future tariffs cover servers and other equipment used in data centers, the administration would have to decide whether similar exemptions remain necessary. Without them, higher equipment costs could increase the expense of building artificial intelligence infrastructure in the United States. With broad exemptions, however, the tariffs could have a smaller immediate effect on imports and therefore provide a weaker incentive for domestic manufacturing.

This creates a policy tradeoff. A tariff needs to be high or broad enough to encourage companies to reconsider their supply chains, but exemptions may be necessary to prevent the policy from increasing costs for industries that the United States is simultaneously trying to expand.

The issue is particularly sensitive because many American technology companies design products domestically but depend on international manufacturing. Increasing import costs does not automatically mean that production will move to the United States. Companies may instead change suppliers, absorb some costs, raise prices or shift manufacturing to another country that remains outside the highest tariff category.

China and Artificial Intelligence Add Another Layer

The semiconductor tariff debate is also connected to the US competition with China over artificial intelligence. Washington has already imposed extensive export controls designed to restrict China’s access to advanced computing technology. At the same time, Chinese technology companies have sought alternative ways to obtain computing capacity, including through overseas cloud services and other parts of the international technology system.

That makes control of the semiconductor supply chain more complicated than simply reducing imports. The United States is trying to achieve two related objectives: maintain an advantage in advanced computing while increasing domestic production of strategically important technologies.

The January tariff action included specific advanced computing chips from companies such as Nvidia and AMD. The administration described the measures as part of a broader effort to reduce reliance on foreign manufacturing and strengthen domestic semiconductor capacity.

The reported new proposal could therefore be viewed as another attempt to close gaps between semiconductor policy, industrial policy and artificial intelligence strategy. If computers, servers and other devices containing advanced chips become subject to tariffs, the policy could cover more of the infrastructure through which computing power reaches businesses and consumers.

However, that broader approach also increases the possibility of unintended effects. Artificial intelligence companies depend on large quantities of imported hardware, while American consumers and businesses purchase technology assembled through international supply chains. Higher import costs could eventually be reflected in prices if companies cannot absorb them.

The impact would depend heavily on the final tariff rate, the products covered and the exemptions provided. Because the proposal remains under discussion, it would be premature to assume that every technology product would face the same treatment.

The Main Test Will Be Whether Tariffs Produce New Capacity

The most important measure of the policy will ultimately be whether it produces additional US semiconductor capacity rather than simply increasing the cost of imported technology. The administration has already linked semiconductor tariffs to domestic manufacturing objectives, and the reported proposal could strengthen that connection by tying tariff relief to investment commitments.

That approach could encourage foreign manufacturers to expand their American operations. Companies deciding where to build new facilities must consider not only production costs but also market access. If tariff-free access to the large US market depends partly on domestic investment, manufacturing inside the country could become more attractive.

The difficulty is that semiconductor production is highly specialised. Different stages of the supply chain are concentrated in different countries, and moving one stage to the United States does not necessarily eliminate dependence on foreign suppliers. A new US fabrication plant may still rely on imported equipment, materials, advanced packaging or components.

The administration has already applied similar thinking to other strategic materials. In August, the White House introduced measures covering polysilicon and related products, combining minimum import prices, tariffs and incentives linked to domestic investment. The approach indicates a broader policy preference for using trade restrictions alongside incentives to encourage domestic production.

For semiconductors, the stakes are higher because chips are embedded throughout the modern economy. A policy that successfully encourages new American production could strengthen supply security over time. A policy that mainly increases the cost of imported components could instead raise expenses for manufacturers and technology users without producing capacity quickly enough to offset those costs.

The reported semiconductor tariff proposal therefore remains less a finished policy than an indication of where US trade strategy may be heading. The administration appears to be considering a system in which access to the American market is increasingly connected to investment in American manufacturing. Whether that approach works will depend on the final tariff structure, the treatment of exemptions, the willingness of foreign companies to build facilities in the United States and the speed with which those investments can become productive.

For the technology industry, the immediate uncertainty is likely to remain around cost and supply chains. For Washington, the larger test is whether tariffs can encourage enough domestic production to justify the additional costs they may impose on a sector that is increasingly central to artificial intelligence, computing and the wider US economy.

(Adapted from LiveMint.com)

Categories: Economy & Finance, Geopolitics, Regulations & Legal, Strategy