AI industry says Trump chip tariffs could "doom" innovation

The Trump administration may be preparing to announce sweeping new semiconductor tariffs at what the tech industry considers the worst possible time. Politico reported this week, citing about eight people granted anonymity to discuss the administration's plans, that a wide range of new tariffs on chips — which the industry fears could "doom" American artificial intelligence innovation — could be imposed within "weeks or months." According to Politico's sources, cited by Ars Technica, the framework under consideration could "dramatically expand the number of tech products subject to the duties, hitting not just chips but potentially many of the goods made with them, such as gaming consoles or the servers that fill data centers." South Korean outlet AI Times, corroborating the report separately, specified that the products under review include laptops, gaming consoles and the servers used in AI data centers.

A price tag in the tens of billions

Trade groups have warned since Trump took office that taxing both semiconductors and the downstream products built with them — potentially even used or refurbished goods containing chips — would be economically ruinous. The Computer and Communications Industry Association (CCIA) estimated in June that this kind of broad tariff approach would cost the US economy about $90 billion a year in lost GDP and cause roughly 20% of the data center projects planned through 2030 to be delayed or cancelled outright. The CCIA even warned the tariffs could push more data center construction outside the US altogether — the opposite of what the policy is meant to achieve. In a May letter to Treasury Secretary Scott Bessent, co-signed by about 20 trade groups, the CCIA laid out further ripple effects if products aren't exempted from the tariffs.

The industry's warning to consumers

That letter argued that prices of "everyday tools" — smartphones, laptops, tablets, smartwatches, connected devices and vehicles — could rise at a time when US households are already budget-strained, and that tariffs could delay the launch of new devices, including ones featuring the latest AI technology. "Consumer devices are the primary interface through which Americans access AI-powered tools," the letter said. "AI only delivers on its promise when people can actually use it — and tariffs that price consumers out of the device market would slow AI adoption at the very moment the United States is positioned to lead." To shield AI firms, the administration is weighing some tariff relief, but according to Politico's sources it would likely be tied to foreign chipmakers' investment in US manufacturing — the approach reportedly favored by Commerce Secretary Howard Lutnick.

• Estimated annual GDP cost of the broader tariff approach, per the CCIA: about $90 billion.
• Share of data center projects planned through 2030 that could be delayed or cancelled: roughly 20%.
• Global semiconductor revenue forecast for 2026, per Gartner, driven by shortage-fueled price increases: $1.6 trillion.
• TSMC's committed investment in Arizona chip-manufacturing facilities: $265 billion — yet, per AI Times, the US is projected to reach only about 30% of cutting-edge chip production capacity even once that investment is fully operational.
• Estimated minimum time needed to build adequate domestic advanced-chip capacity, per industry observers cited by AI Times: five years.

Why the reshoring math doesn't add up, according to critics

Lutnick reportedly wants to apply tariffs broadly to prioritize rebuilding the domestic chip supply chain, and four sources told Politico he plans to let a set volume of chips enter the US duty-free, sized to how much companies pledge to produce on American soil — with different rates possibly applied by country. Critics say that structure would leave a large gap between the tariff-free supply and what the industry actually needs: "the volume they're talking about granting duty-free wouldn't cover the hyperscalers alone, let alone the rest of the industry," one industry representative told Politico. "Those are chips we physically can't buy here, because the capacity doesn't exist yet. The math literally just does not work." A tech official who served in the first Trump administration put it more bluntly: "This may be the single dumbest way imaginable to pursue American dominance in AI. It's like kneecapping yourself at the starting line." Talks between the industry and Commerce officials have reportedly "trended in a negative direction" in recent weeks, even as lobbying has intensified; AI Times reports that companies are specifically pushing to keep the exemptions for data centers, R&D, startups and consumer products that were granted under the administration's initial 25% tariff on advanced AI accelerators announced in January.

For a UK business that relies on AI hardware — cloud infrastructure bought from US-based providers, laptops, or the servers that power data centers — this dispute matters well beyond American borders. Any tariff that raises the price of chips, or of the finished machines built with them, ripples through global supply chains: equipment sourced from US vendors, or built with US-supplied semiconductors, could get pricier long before Washington finalises the policy. The scale of the numbers at stake, and how sharply US officials themselves disagree over the reshoring math, suggest this fight is far from settled — and its outcome will help set the price of AI hardware well beyond the United States.