The ongoing debate surrounding semiconductor export controls is less about trading goods and more about leadership in artificial intelligence.
The Trump administration’s move to restrict advanced U.S. semiconductors from reaching Chinese companies outside mainland China seems like a narrow trade issue. However, it reflects a broader reality where the U.S.-China competition hinges significantly on artificial intelligence.
This raises a fundamental question: Should Chinese firms be allowed to acquire advanced U.S. chips via foreign subsidiaries while still being governed by Chinese law?
This isn’t just a theoretical issue. China’s National Intelligence Law from 2017 mandates that organizations and citizens support national intelligence operations as required by law.
John Moolener, who leads the House Select Committee on the Communist Party of China, pointed out that a company’s location doesn’t exempt it from Chinese legal obligations. Additionally, there are concerns over whether parent companies remain answerable to authorities in Beijing.
What may appear to be a technical debate about chips is actually much more complex.
Advanced semiconductors are crucial for modern AI, enabling everything from training models to analyzing intelligence, processing vast data sets, and conducting scientific research. Access to these technologies has become key to achieving technological dominance.
China has dedicated considerable resources to reducing its reliance on foreign semiconductor supply chains, notably through the National Integrated Circuit Industry Investment Fund, known as the Big Fund, which has channeled billions into domestic chip innovation. Recently, China initiated a new phase with an investment of approximately $47.5 billion.
Meanwhile, U.S. authorities have reported ongoing efforts to evade these export controls. The Department of Justice has taken action against illegal technology transfers to China, including a case in January 2026 where federal agents dismantled smuggling rings that procured Nvidia GPUs. These chips were repackaged under a false company name before being sent to China. In a separate incident, three individuals faced charges for attempting to ship around $170 million worth of restricted chips through Thailand to obscure their destination.
These incidents underline persistent incentives for stakeholders to bypass restrictions rather than being isolated events.
Relying solely on export controls isn’t sufficient. AI capabilities hinge on various factors beyond just chips, including cloud infrastructure, available data, software artistry, engineering expertise, and advanced algorithms. Tightening one aspect while neglecting others might only shift the problem rather than resolve it.
Chairman Moolener’s SCALE Act aims to create a more robust framework for semiconductor limits by establishing clearer criteria for their application. While opinions may differ on its specifics, the underlying issues resonate widely.
The effectiveness of export controls is directly tied to the gaps they create. When policymakers recognize the weaknesses in their strategies, the focus shifts to enforcement and maintaining credibility.
The United States still possesses significant strengths, with its universities, research institutions, capital markets, and tech firms among the world’s leaders. However, these advantages depend on supportive policies that foster innovation while safeguarding technologies impacted by national security concerns.
The global race for AI dominance unfolds across research labs, cloud centers, universities, startups, and government agencies. The chips circulating through these channels today will play a critical role in determining who maintains a technological lead a decade from now.
Effective export controls go beyond mere trade disputes; they aim to protect crucial advancements while still manageable.
