The race for artificial intelligence leadership will be the defining economic and national security competition of the 21st century. President Donald Trump understands this, which is why the delegation on his recent trip to China included a who’s who of tech titans. It is a telling sign of just how central AI technology has become to the highest levels of geopolitical deal-making.
For now, the United States holds a meaningful lead. But that advantage is not guaranteed. It depends on whether the rest of the world continues to build on American technology or begins to adopt an alternative stack. That question sits at the center of a growing debate in Washington.
A number of analysts from across the political spectrum have argued that the best way to preserve U.S. leadership is to further restrict China’s access to American AI chips that are a generation or two behind our most advanced chips. A recent op-ed in the New York Times, for example, called for tightening controls on “critical technologies.” Meanwhile, another piece in the Atlantic warned that lifting export controls throws away America’s AI dominance.
The instinct behind these arguments is understandable. AI systems are advancing rapidly, with clear implications for cybersecurity and national defense. Thoughtful guardrails are not just appropriate; they are necessary. But broad, overly restrictive controls on chip exports are the wrong way to secure American leadership, and the existence of risk does not justify treating every commercial export as a national security emergency.
After the United States moved to restrict advanced semiconductor exports in 2022, China accelerated its push toward self-sufficiency. Backed by state investment and national urgency, Chinese firms expanded rapidly across the AI stack. Now, Huawei’s Ascend chips, once viewed as years behind Nvidia, are rapidly improving in both computing capability and scalability. Its latest generation already powers major AI training systems in China, and the company has openly stated its ambition to compete with Nvidia’s Hopper-, Blackwell-, and Rubin-era chips over the coming years.
At the same time, Chinese developers adapted in ways policymakers did not anticipate. Faced with limited access to cutting-edge hardware, they focused on efficiency — building models that could deliver competitive performance using fewer resources. That shift has proven consequential. By 2025, Chinese organizations had released more than 1,500 large language models, accounting for roughly 40% of the global total, with many gaining traction across emerging markets.
This points to a deeper flaw in the current debate. It treats AI competition as a static race for compute, when in reality it is a dynamic contest over ecosystems. The defining question is not simply who has the most powerful chips. It is whose technology becomes the foundation on which the world builds.
When U.S. firms are excluded from major markets, competitors do not pause; they fill the gap. Chinese companies are already offering integrated AI solutions abroad, bundling hardware, cloud services, and software into accessible, cost-effective packages. In fact, Chinese cloud providers have committed roughly $84 billion in AI infrastructure investment through 2027. Often, these cloud agreements and hardware come at a much lower cost than U.S. alternatives, forcing would-be allies onto CCP-owned systems.
Over time, this creates parallel technology ecosystems operating outside American influence. Once established, those ecosystems are difficult to displace. Developers build tools around them. Governments procure based on them. Data flows reinforce them. What begins as a short-term export restriction can evolve into a long-term loss of market share and, with it, strategic leverage.
There is also a direct cost at home. Companies depend on global sales to finance research, development, and next-generation manufacturing. Cutting off access to large markets shrinks the capital base that sustains U.S. innovation, weakening the very advantage policymakers seek to protect.
None of this suggests that export controls should be abandoned. Targeted restrictions on the most sensitive technologies remain an essential national security tool. But there is a meaningful difference between protecting a narrow edge of capability and broadly walling off commercial technologies altogether.
TRUMP ANNOUNCES NEW APPLE AND INTEL PARTNERSHIP TO BOOST CHIP PRODUCTION IN THE US
Recent policy adjustments from the Trump administration that allow the sale of advanced AI chips to vetted buyers while holding frontier technologies in reserve reflect a step in the right direction. By keeping U.S. firms engaged in global markets while safeguarding the most sensitive capabilities, this approach preserves both security and competitiveness. The alternative — doubling down on sweeping restrictions — risks accelerating exactly what policymakers hope to avoid: a self-sufficient Chinese AI ecosystem that competes not just domestically, but globally.
AI leadership will not be determined by who can build the highest wall. It will be determined by whose technology the world chooses to adopt. If the United States wants to hold the lead, it must remain present in the markets, systems, and partnerships that will define the next generation of AI. Retreating from those arenas does not preserve advantage. It concedes it.
Ron Wright is the co-founder of the National Artificial Intelligence Association (NAIA) America’s largest business coalition, representing companies and their allies, building in AI.