What happens when a superpower taxes its own path to artificial intelligence?

Sejal Sharma is IE’s AI columnist, offering deep dives into the world of artificial intelligence and its transformative impact across industries. Her bi-monthly AI Logs column explores the latest trends, breakthroughs, and ethical dilemmas in AI, delivering expert analysis and fresh insights. To stay informed, subscribe to our AI Logs newsletter for exclusive content.

Last week, the world watched with raised eyebrows as the US and China engaged in a game of tariff ping-pong. Keeping up with who’s levying how much tariff on whom has been difficult as we move into the thick of the tariff drama. To be clear, things as they stand today (and they can change very quickly) are that the US has imposed an eyewatering 145 percent tariff on Chinese goods, and China has imposed a 125 percent tariff on American goods.

Things are still unclear in terms of how this impacts the development of artificial intelligence-related technologies in the US. The Trump administration hinted at many moving parts, like the looming tariffs on semiconductors on April 15.

Trump had initially, on April 2, announced an exemption for semiconductors. Moreover, US customs notice on April 11 said smartphones, computers, and other electronics would be excluded from the 125 percent tariff on Chinese imports, which are about a quarter of US imports from China. 

However, on Sunday, Trump contradicted this on social media, saying there was no exemption for these products. He posted on Truth Social, “We are taking a look at Semiconductors and the WHOLE ELECTRONICS SUPPLY CHAIN in the upcoming National Security Tariff Investigations.”

And on April 15, the Trump administration made a move that could lead to new semiconductor tariffs. In the afternoon, government notices popped up online saying they’ve kicked off national security investigations into imported chips. The probe isn’t just restricted to the final products, aka chips. It also looks at the machines used to make these chips and anything with chips in it.

Tariffs on chips may slow the US’ AI ambitions

Semiconductors are small, powerful pieces of tech that are the basis of modern electronic equipment. They are used on everyday devices such as mobile phones, refrigerators, laptops, and other electronic items. They are also used to build powerful AI machines.

These AI chips are essential for the US to win the AI race it is embroiled in with China and other nations. Control over semiconductor production and access to cutting-edge designs has become a central concern in the global race for AI dominance.

The chips are the engines that power large-scale machine-learning models. From training generative AI systems like ChatGPT to running facial recognition in smartphones or autonomous driving algorithms in electric vehicles, these chips are designed to perform parallel computations at scale. Without them, modern AI would be sluggish and inefficient, and it would be simply impossible to achieve artificial general intelligence (AGI), which the big tech US companies are hurling towards.

Trump’s tariffs would mean an increased cost to import chips from nations like Taiwan and South Korea, which the US currently sees as a national security risk owing to its reliance on these countries.

But why? The idea behind Trump levying these tariffs worldwide, not just on China, is to push manufacturing goods like semiconductors in the US. President Trump has taken a hardline approach to making the US a manufacturing hub. He’s all about bringing jobs back home, cutting reliance on foreign supply chains, and reshoring industries that he sees as crucial. Like semiconductors, steel, and pharmaceuticals. The ‘Make in America’ approach is now synonymous with his patriotic branding of ‘Make America Great Again.’

And companies might already be falling in line? Nvidia, for example, announced on April 14 its plans to manufacture AI chips and supercomputers entirely within the US.

“Together with leading manufacturing partners, the company has commissioned more than a million square feet of manufacturing space to build and test NVIDIA Blackwell chips in Arizona and AI supercomputers in Texas,” said the company.

The move aims to mitigate the impact of tariffs and align with the US government’s push for domestic production. 

“Within the next four years, NVIDIA plans to produce up to half a trillion dollars of AI infrastructure in the United States through partnerships with TSMC, Foxconn, Wistron, Amkor and SPIL,” added the company.

Similarly, other tech giants are exploring localizing their supply chains, reducing dependency on foreign components, and investing in domestic manufacturing capabilities. 

But this is going to take time. 

Destabilizing America’s supply chains?

Trump’s tariffs will also be targeting data center equipment. China is still a major player, especially regarding rare Earth processing, electronic components, and low-cost computing hardware. These tariffs threaten AI development by making data center construction and operation significantly more expensive. 

Due to global supply chain disruptions and long waitlists, high-end GPUs and backup generators are now harder to source affordably. The US heavily imports HVAC systems and components from Mexico and China, which are used for cooling purposes in data centers. 

In all likelihood, the US will also have to rebuild its energy grids to meet the growing electricity demand of data centers. While the electricity is generated domestically, the tools to generate and store it are often imported.

This comes at a time when companies like Google and Microsoft plan to spend $75 billion and $80 billion, respectively, on AI data centers this year alone. They are likely to be impacted as well.

How tariffs could skew the US-China AI race

Earlier this year, Trump’s Oval Office included announcements from OpenAI, Oracle, and Japan’s SoftBank of a proposed $500 billion AI infrastructure initiative dubbed “Stargate.” He said that day, “I’m going to help a lot.” After the ‘Liberation Day’ tariff announcement, Oracle’s and Nvidia’s stock plunged, and other big tech companies’ stock prices were taking a nose dive. Some rebounded after Trump announced a 90-day halt for most countries, except China.

However, Trump’s tariff regime, which many believe may have been concocted using generative AI’s help, may delay or divert that investment. 

Industry experts warn that rising costs and uncertainty could push US firms to build AI infrastructure abroad in countries with more stable trade environments. 

This could distort the competitive balance between the US and China in the AI race. China, meanwhile, has been accelerating its efforts to build homegrown alternatives, whether through national chip projects or massive investments in local fabs. With limited access to Nvidia’s high-end AI chips due to earlier US export restrictions, Chinese firms like Huawei, Baidu, and Alibaba are racing to develop domestic equivalents. For example, look at DeepSeek, a low-cost AI model built using less efficient Nvidia chips.

These tariffs may accelerate what many analysts call the “AI decoupling.” Countries are being forced to rethink their reliance on either Washington or Beijing. Europe, India, and the Gulf states are exploring how to build and host their own AI ecosystems, not just for strategic autonomy but because it’s increasingly too risky or expensive to depend on a global supply chain that could collapse with the next tariff tweet.

Trump’s “Manufacture in America” pitch sounds bold, but the ground reality differs. Setting up domestic chip fabrication plants, securing raw materials, and training the necessary talent are decades long. 

To illustrate this point, a report of a dinner conversation between former US president Barack Obama and Apple founder Steve Jobs has been doing the rounds. During dinner, Obama asked Jobs why the US could not produce iPhones back home. Jobs reportedly replied that “those jobs aren’t coming back,” pointing to overseas factories’ unmatched scale, speed, and flexibility, especially in China. 

According to the report, Apple executives at the time believed that the sheer number of overseas factories and the flexibility, diligence, and industrial skills of foreign workers far outpaced what American manufacturers could offer.

And the US stopped manufacturing a long time ago.

Even with big subsidies and CHIPS Act incentives, the US is nowhere near its capacity to become self-reliant in semiconductor manufacturing. The costs will be passed on to American companies in the short term. And eventually, to consumers.

What we’re witnessing isn’t just a trade war. It’s a tectonic shift in how nations think about AI sovereignty. The global AI supply chain might fracture into regional blocs: US-led, China-led, India-led, and a growing “non-aligned” group of nations looking to chart their course.