In an era increasingly defined by great-power competition with China, the nation with the strongest commercial technological base will hold a decisive military and economic advantage. Winning the global AI race is therefore not simply an economic objective — it is a national-security imperative.
Thanks to unmatched innovative capacity, the United States still leads in artificial intelligence. But maintaining that lead requires more than protecting today’s breakthroughs. It requires ensuring that tomorrow’s global AI ecosystem — from chips and cloud platforms to software frameworks and development standards — continues to run on American technology.
That is why export controls on AI chips and related technologies present such a complex policy challenge. Used judiciously, they can protect sensitive capabilities. Applied too broadly, they risk producing the opposite of their intended effect.
History suggests that sweeping technology restrictions often accelerate indigenous innovation in the targeted country. Faced with limited access to U.S. semiconductors, Beijing has doubled down on domestic chip fabrication, software optimization, and alternative AI development strategies. Rather than freezing China’s progress, restrictions have encouraged the emergence of parallel supply chains and technological ecosystems increasingly outside American influence.
This shift carries real strategic consequences. When global developers build on American chips and platforms, the United States helps shape standards, governance norms, and technical architectures. When they migrate elsewhere, that influence erodes.
China’s long-standing ambition to become a technology superpower makes this dynamic particularly consequential. For decades, its progress depended significantly on access to U.S. innovation. Export controls have begun altering that balance, pushing Chinese firms toward self-sufficiency while insulating domestic champions from foreign competition. The risk is not merely commercial loss; it is the gradual fragmentation of the global AI ecosystem.
At the same time, constrained access to advanced chips has driven Chinese developers to emphasize efficiency — optimizing software, refining model architectures, and extracting more performance from less computing power. Some Chinese open-source AI models have begun narrowing performance gaps with American systems, illustrating how hardware restrictions can inadvertently spur innovation elsewhere while domestic chip ecosystems mature behind protected markets.
This leaves U.S. policymakers at an inflection point.
Despite these realities, Congress is today considering proposals such as the AI Overwatch Act (H.R. 6875) that would layer additional statutory restrictions and congressional approval requirements onto advanced AI chip exports. However well intentioned, such measures risk hardening the very kind of overbroad export regime that has already encouraged the emergence of rival technological ecosystems beyond American influence.
The recent policy shift under the Trump administration reflects a more calibrated approach. Allowing sales of certain advanced AI chips to vetted commercial customers in China — while continuing to restrict the most cutting-edge hardware and frontier capabilities — aims to preserve both security and strategic leverage.
The logic is straightforward: keeping American firms engaged in major global markets helps sustain revenue for research and development, prevents foreign competitors from gaining captive markets, and keeps global AI infrastructure anchored to U.S. technology standards. It’s for reasons such as this that David Sacks, Chair of the Chair of the President's Council of Advisors on Science and Technology, tweeted his agreement that measures such as the AI Overwatch Act would “take away President Trump’s authority as Commander in Chief and undermine his America First strategy.”
China remains the world’s largest semiconductor consumer. Losing sustained access to that market has already cost U.S. companies tens of billions of dollars in potential revenue that might otherwise have funded next-generation innovation. Equally important, once developers rebuild software ecosystems around alternative chips, they rarely revert. Maintaining market presence therefore has long-term strategic implications.
As history has shown, dependence can also be a form of leverage in and of itself. The dominance of the U.S. dollar in global finance illustrates how widespread reliance on American systems creates enduring influence. A similar dynamic exists in technology. When competitors and rivals rely on American chips, software, and infrastructure, Washington retains visibility, influence, and the ability to shape global standards. Complete technological decoupling, by contrast, risks forfeiting those advantages.
That leverage has direct national-security relevance. Much of the case for broad export controls rests on the assumption that denying China access to advanced AI chips meaningfully constrains military capabilities. In practice, however, many operational defense applications rely less on frontier training hardware than on optimized models, specialized systems, and integration across platforms. As China’s domestic ecosystem continues to improve, the marginal security benefit of sweeping commercial restrictions diminishes. Maintaining selective technological dependence, by contrast, preserves visibility, influence, and standard-setting power — advantages that are forfeited when competitors are pushed toward full technological separation.
We will not win the AI race against China by denying access but rather by maintaining superiority — faster innovation cycles, deeper capital markets, stronger research ecosystems, and wider global adoption of American platforms. Engagement, when carefully structured, reinforces those advantages.
The central policy question is therefore not whether export controls should exist, but how they should be designed. Broad restrictions risk accelerating self-sufficiency abroad. Targeted controls — combined with sustained commercial engagement — preserve leverage while protecting core national-security interests.
In a prolonged strategic competition, staying ahead matters more than trying to hold others back. Smart export policy should reinforce that advantage — not inadvertently erode it.
