The Trump administration is considering a sweeping new round of tariffs on semiconductors, a move that is already unsettling the U.S. technology sector as companies struggle to secure enough advanced chips to meet the explosive demand created by artificial-intelligence data centers.

One proposal being examined would significantly broaden the range of technology products covered by the tariffs. The duties could extend beyond semiconductors to products that depend on them, including laptops, gaming consoles and the servers packed into data centers. Commerce Secretary Howard Lutnick supports an approach that would link foreign companies’ access to tariff relief to commitments to expand semiconductor manufacturing in the United States, according to four people familiar with the discussions.

Officials are also considering gradually introducing the new tariffs, according to those people. However, the proposal remains under development and could change substantially before the administration settles on a final framework.

The individuals familiar with the discussions were granted anonymity because they were describing private government deliberations.

The prospect of additional chip tariffs is particularly concerning for U.S. technology companies because the industry is already facing tight supplies of advanced semiconductors. Demand has surged as companies pour money into the construction of data centers designed to support artificial-intelligence systems.

Industry advocates say they support President Trump’s objective of expanding semiconductor manufacturing inside the United States. But they point out that advanced chip factories require billions of dollars and can take many years — potentially decades — to construct and bring fully online. In the meantime, American companies remain highly reliant on Asian suppliers, including Malaysia, South Korea and Taiwan. Taiwan alone accounts for more than 90 percent of the world’s most advanced semiconductor production.

The dispute highlights a fundamental tension within the administration’s technology strategy. Efforts to rebuild America’s semiconductor industry could collide with Trump’s ambition to make the United States the global leader in artificial intelligence. While tariffs are intended to encourage manufacturers to move production to the United States, technology companies warn that taxing imported chips could make the hardware needed for the AI boom more expensive and potentially force companies to reconsider planned data-center projects.

“This data center buildout, in scale and dollars, has been compared to building the transcontinental railroad,” said Jonathan McHale, digital policy chief at the Computer and Communications Industry Association, which counts major tech companies such as Amazon, Google and Meta as members. “Anytime you add to the cost and decrease predictability you make it more difficult to invest, and you are putting that in jeopardy.”

The White House defended the administration’s use of trade policy as a way to strengthen domestic manufacturing.

“Reshoring semiconductor manufacturing is a top priority for President Trump, whose policies have already secured hundreds of billions of dollars of investments in this key sector,” said White House spokesperson Kush Desai. “The Trump administration remains focused on delivering more investments and economic relief for the American people while safeguarding our national security.”

The Commerce Department did not answer multiple requests for comment.

Technology companies have responded with an aggressive lobbying campaign aimed at persuading the Trump administration to narrow the expected tariffs. Industry representatives want the new measures to more closely resemble duties announced by the president earlier this year, which included broad exemptions for data centers and other domestic applications and therefore limited their effect on the sector.

Industry representatives and lobbyists have also held increasingly frequent meetings with senior administration officials since the beginning of summer. Among those involved is Commerce Undersecretary Jeffrey Kessler, who oversees the department’s Bureau of Industry and Security, as well as Lutnick.

Lobbyists are warning that tariffs could slow the rapid expansion of data centers by making it more difficult for American companies to obtain the enormous quantities of semiconductors required to power the AI industry. The timing is especially concerning because U.S. technology giants are already engaged in a record-breaking spending surge on artificial intelligence, committing hundreds of billions of dollars to massive data-center campuses while competing for scarce, expensive advanced chips.

“This may be the single dumbest way imaginable to pursue American dominance in AI,” said one tech official from a major industry group that also served in the first Trump administration. “It’s like kneecapping yourself at the starting line.”

Several lobbyists came away from their discussions believing Kessler was more receptive to the industry’s concerns. However, they also emphasized that Lutnick has the final say and that Kessler was unlikely to substantially oppose his preferred approach.

“Every conversation is some version of: ‘Please, tread carefully. You don’t understand how much demand is coming, or how little capacity we have to meet it,’” said the tech official.

More recent discussions have become less encouraging for the technology industry, according to three people familiar with the matter. Administration officials have indicated that they may abandon exemptions similar to those included in the January tariffs. Those exemptions previously covered data centers, research and development, startups, consumer applications, civil industrial uses and public-sector projects, along with “other uses” that the Commerce secretary determines would strengthen the U.S. supply chain.

Instead, Commerce officials have increasingly argued in private conversations that the tariffs must apply more broadly to companies in order to pressure semiconductor manufacturers into expanding production within the United States.

Under the framework favored by Lutnick, the United States would establish a duty-free allowance for a specified quantity of imported chips. That allowance would be connected to the amount of semiconductor production companies commit to establishing domestically. The approach could create a widening mismatch between the number of chips eligible for tariff-free entry and the much larger quantities American businesses actually require, according to four people familiar with the proposal.

People briefed on the discussions said Commerce officials have not yet determined the exact tariff rates or several other important details. One proposal would establish different tariff rates and import quotas for individual countries, with separate guidance applying to major semiconductor manufacturers in each nation, according to four people familiar with the matter.

Most semiconductors and related components used by American companies come from Asian trading partners. Beyond the goal of expanding advanced chip manufacturing in the United States, administration trade hawks are also concerned about the national-security risks created by U.S. dependence on Taiwanese semiconductor production in the event of a Chinese invasion of Taiwan.

“Getting supply chains de-risked is one of the central geopolitical questions of our lifetime,” said Michael Sobolik, a senior fellow at the right-leaning Hudson Institute think tank and a former aide to Sen. Ted Cruz (R-Texas). But Sobolik added, “It’s going to be really expensive to, at scale, build a lot of these chips in the United States, because there are a lot of other cheaper places in the world where you can do that.”

Technology lobbyists and economists say the consequences could extend well beyond data-center construction. Tariffs would increase the price of imported semiconductors used by U.S. companies to manufacture servers, computers, televisions and other electronics. The measures could also hurt American chip designers such as Nvidia and Advanced Micro Devices, which depend on overseas manufacturers to produce their semiconductors.

The tariffs could also put companies such as Apple at a competitive disadvantage in overseas markets. Foreign manufacturers could purchase identical chips without paying the U.S. duties, while semiconductor suppliers in allied countries might have greater incentives to pursue customers in China.

Sujai Shivakumar, an economist who leads the Renewing American Innovation program at the Center for Strategic and International Studies in Washington, said tariffs alone cannot solve the United States’ semiconductor-production problems. Building a competitive domestic industry requires far more than making imported chips more expensive.

“Higher tariffs can change relative prices, but in themselves, they don’t produce more skilled technicians,” Shivakumar said. “They don’t shorten the permitting timelines. They don’t expand reliable power and water infrastructure. They don’t create qualified suppliers.”

One technology-sector representative emphasized that the industry does support expanding semiconductor manufacturing in the United States.

“The data center industry would love to have a strong domestic semiconductor manufacturing industry, and we’re here to support that on-shoring,” the person said. “But the volume and scale of manufacturing of advanced node chips and memory chips is just not where it needs to be.”

The proposal to connect tariff exemptions with domestic investment resembles a trade agreement Trump reached with Taiwan in January. That arrangement allows Taiwan Semiconductor Manufacturing Co., which manufactures chips designed by U.S. companies such as Apple and Nvidia, to bring certain chips into the United States without tariffs.

For now, however, that arrangement covers only a limited portion of the advanced chips required by the market. TSMC has pledged $265 billion toward its Arizona facilities, representing the largest foreign direct investment in U.S. history. Even after the facilities are fully built, however, the company expects only about 30 percent of its most advanced manufacturing capacity to be located in Arizona, with much of that capacity still years away.

One person from the trade and technology sectors who regularly meets with senior Trump administration officials estimated that establishing sufficient domestic chip-manufacturing capacity would take more than five years. That timeline is substantially longer than the phase-in periods Trump has used for previous tariffs, suggesting that major technology companies could face higher costs long before U.S. production can fill the gap.

“The math literally just does not work,” said the tech representative. “The volume they’re talking about granting duty-free wouldn’t cover the hyperscalers alone, let alone the rest of the industry. Those are chips we physically can’t buy here, because the capacity doesn’t exist yet.”