This piece was co-authored by Pablo Garcia Quint, a Technology and Innovation Policy Fellow at Libertas Institute
In the race to lead in AI, the worst-case scenario for the U.S. is to give our foreign adversaries an unnecessary competitive edge. But that’s exactly what new tariffs threaten to do. Despite one of the Administration’s first executive orders sending the strong message that the U.S. is all-in on AI, steep import taxes set to resume in 90 days sends a new, mixed signal to global investors— we want to lead the world in AI, but only if we can build it in a bubble.
Here’s the problem — AI development isn’t just about stamping “Made in America” on a server rack. It’s about letting the best ideas win, no matter where they come from. Whether it’s Palo Alto or Taiwan, the future of AI is borderless because capitalizing on ideas and skills from abroad also benefits the progress we make at home by complementing the things we lack. Tariffs, meant to spur domestic production, risk kneecapping the very companies trying to bring next-gen intelligence to life. If foreign firms see the U.S. as expensive and hostile, they’ll go elsewhere. That’s not just protectionism, it’s self-sabotage.
Consider NVIDIA, the chip manufacturer powering the AI boom. It designs its chips in the U.S. but manufactures them in Taiwan. Those chips power everything from LLMs such as ChatGPT to Google’s quantum devices. While the recent tariff exemptions for the semiconductor industry offer some reassurance, they don’t solve the bigger issue: What happens when the most important chip designer in the world, NVIDIA, can’t rely on a secure supply chain? We saw some of the effects back in 2021, with an increasing demand accompanied by massive shortages in supply. And now, thanks to new U.S. export controls, NVIDIA needs a license to sell its latest H20 chips to China, risking billions in lost revenue.
But, the latest developments show that the tariffs are a negotiation point, with the idea of reducing our foreign dependence to boost domestic production. What makes this potentially self-sabotaging is that the U.S. risks damaging international relations, which will create an artificial shortage and put rival tech companies from our foreign adversaries ahead.
This uncertainty doesn’t just threaten production timelines, it affects developers, disrupts innovation, and puts the broader economy on edge. So the broader question becomes: as tariffs make imports more expensive, what real incentives are we offering companies to build at home to offset those tariff-related costs?
Even if the U.S. were to subsidize production, prior experience has shown lackluster results. A prime example is the CHIPS Act of 2022 —$52.7 billion thrown at semiconductor manufacturing and research to help domestic players like Intel stay competitive and attract foreign companies like TSMC, Samsung, and Micron. However, the CHIPS Act subsidies failed to pan out because subsidies can’t replace real market demand.
Tariffs — just like subsidies — put the wrong incentives on businesses to game the system instead of encouraging them to create a superior product. In Utah alone, the latest tariff increases are expected to add $3.1 billion to the cost of imported goods, a burden that will slow future AI development by driving up the price of key inputs needed for upcoming data centers. Even with market-based solutions like Tarifflo — a Utah-based startup that streamlines the complex import documentation process across different regulations and tariffs — the broader costs of tariffs are unlikely to be meaningfully offset.
If the U.S. wants to stay at the forefront of AI, the way forward isn’t tariffs; it’s not subsidies; and it’s not streamlined import-export documentation. It’s giving builders the freedom and certainty to build, and capitalizing on that freedom by leveraging our comparative advantages from our place in the global economy. Tariffs won’t revive the golden age of American manufacturing—they’ll only delay the golden future of AI. Instead of building walls around our economy, we should be building bridges: to talent, to supply chains, and to global partners who want to shape the next technological revolution with us.
