When Markets Meet National Security
Washington faces an increasingly uncomfortable tension between two legitimate priorities: protecting American commercial interests and preventing the transfer of advanced military technology to strategic competitors. Congressional inquiries into semiconductor export compliance illustrate just how thorny that balance has become.
The companies at the center of these debates have become central to artificial intelligence advancement. Advanced graphics processing units power some of the world's most sophisticated computing applications, from cloud services to defense systems. That same computational muscle, however, has made semiconductor manufacturers natural focal points for policymakers worried about technological advantage in great-power competition.
The Export Control Framework
The United States has long maintained restrictions on the export of advanced semiconductors and manufacturing equipment to China and other nations deemed security risks. These controls aim to prevent technologies with military applications from reaching adversaries. The logic is straightforward: if advanced chips power both artificial intelligence and weapons systems, restricting their flow protects American strategic superiority.
But implementation is far messier than the principle. Leading semiconductor chips serve dual purposes—they enable both civilian artificial intelligence applications and military-grade computing. Drawing precise lines between what can and cannot be sold requires regulators to make technical judgments about capability and intent, judgments that are notoriously difficult to make with certainty and that shift as technology evolves.
Industry participants have maintained that they comply with all export regulations as written. They have also pointed out, with some validity, that designing and marketing products specifically to avoid regulatory scrutiny creates its own set of perverse incentives. If American companies simply stop selling to legitimate civilian customers in certain regions, those customers may turn to competitors—potentially foreign rivals whose products the U.S. has less insight into and less ability to influence.
The Legitimate Questions
Congress is right to press for clarity on these matters. Oversight of sensitive export policy is not partisan impulse; it is institutional responsibility. Lawmakers have a duty to understand how effectively export controls are being enforced and whether regulations are achieving their stated security goals or merely shifting market share.
Several substantive questions deserve serious answers. First: Are current regulations actually preventing technology transfer, or are they simply redirecting purchases through intermediaries or competing suppliers? Second: Do companies' marketing practices or product design choices effectively circumvent the letter of the law while respecting its spirit? Third: What insight do regulators actually have into whether exported equipment ends up in civilian applications or in military research facilities?
These are technical and policy questions, not accusations. They deserve testimony, documentation, and straightforward answers.
The Deeper Policy Problem
Yet beneath these specific controversies lies a more fundamental challenge for American policy. The United States has constructed an export control regime based on assumptions that no longer hold perfectly. We assume we can identify which technologies are strategically sensitive and which are not. We assume we can distinguish civilian from military applications. We assume that restricting sales to one nation while allowing them elsewhere will effectively slow an adversary's advancement.
None of these assumptions is entirely false, but all of them have become more complicated in an era of dual-use technology, global supply chains, and rapid innovation cycles. Semiconductors are especially vexing because nearly all advanced chips have both civilian and military applications. The boundary between consumer and military-grade computing has blurred.
A serious export control policy needs to reckon with this complexity honestly. That means acknowledging tradeoffs: restrictions that are too tight might push innovation offshore or spur the development of domestic alternatives in competitor nations. Restrictions that are too loose fail their security purpose. Finding the right calibration requires constant reassessment, not just congressional theater.
What Institutions Should Do
Congress should approach these inquiries as regulators and overseers, not as prosecutors looking for scalps. The goal should be understanding whether current policy is working as intended and whether it needs adjustment—not generating headlines by hammering companies for following the rules.
That said, semiconductor manufacturers and other companies with access to sensitive technology should expect scrutiny. They operate in a heavily regulated domain. Transparency about compliance practices, supply chain controls, and end-use verification is not an unreasonable demand. Companies benefit from the American innovation ecosystem, patent protections, and rule of law; they should welcome the opportunity to demonstrate they are stewarding sensitive technology responsibly.
The executive branch, meanwhile, needs to provide clearer guidance about what the rules actually permit and require. Ambiguity in export policy is bad policy. It creates compliance uncertainty for companies and reduces oversight effectiveness for government.
The Larger Point
These disputes, like many technology policy questions, cannot be resolved by simply choosing between commerce and security. American strength rests on both. We need a thriving semiconductor industry and effective export controls. We need companies with global reach and government institutions capable of protecting genuinely sensitive technology.
That balance is achievable, but only if Congress, regulators, and industry approach these questions as genuine problems to be solved rather than opportunities for partisan advantage or corporate defensiveness. Congressional hearings on these matters are a chance to demonstrate that we still have the institutional maturity to make that work.
