U.S. AI dominance rests on infrastructure, not models—and that's harder to challenge

The article opens by reframing the AI competition away from model performance—the focus that has dominated recent headlines following the rise of low-cost systems from Chinese companies—and instead argues that infrastructure is the true arena of competitive advantage. While observers have been fixated on which AI models are faster, more capable, or cheaper, the author contends that as long as the U.S. controls the underlying infrastructure enabling AI ecosystems, American dominance will persist.

The heart of U.S. AI dominance lies in Nvidia, described as the world's most valuable company. Nvidia controls roughly 85% of the global market for graphics processing units (GPUs), the chips required to train frontier AI models. Beyond hardware, Nvidia's ecosystem stretches to include hardware manufacturers like Broadcom and cloud hyperscalers—Amazon Web Services, Google Cloud, and Microsoft Azure—which provide the infrastructure powering foundational AI developers such as Anthropic, OpenAI, Meta, and Alphabet. The critical binding element is CUDA, Nvidia's software layer, which has become the default environment for AI development and creates high switching costs for anyone attempting to shift to competitors. This integrated ecosystem forms what the article calls a new "U.S. AI industrial complex," with extensive ties to America's defense and intelligence establishment through major binding contracts.

The military dimension of this dominance is substantial. The Department of Defense has signed standard operational agreements with major providers—Google, OpenAI, Microsoft, Amazon Web Services, Oracle, and Nvidia—to deploy frontier AI tools onto classified military networks. The Pentagon's FY2027 budget earmarks more than $54 billion for autonomous warfare and drone systems, funding a newly formed Defense Autonomous Warfare Group (DAWG). OpenAI, xAI, and Google have signed binding contracts allowing the Pentagon 'all lawful use' for defense-related purposes, including autonomous weapons and mass surveillance. Anthropic, despite ongoing litigation against the U.S. government over objections to how its models may be used, reportedly has embedded engineers inside the National Security Agency to adapt its Mythos model for offensive cyber operations. The scale of the commercial-military symbiosis is described as "almost incomprehensible"—ten of the world's largest companies by market cap are American tech firms comprising this ecosystem, representing commercial concentration in the trillions of dollars. Federal AI contracting surged $90.7 billion in 2026 alone, according to the Brookings Institution.

Underwater fiber-optic cables represent a second critical layer of infrastructure dominance. As American cloud hyperscalers expand capacity around the world, they are increasingly building privately owned subsea fiber-optic networks. Meta plans to build an around-the-world subsea cable covering 40,000 kilometers with a projected cost of $10 billion. Google, through its Pacific Connect Initiative, will spend over $1 billion to connect Japan to the South Pacific. These vital data pipelines, owned and operated by Silicon Valley tech giants, account for 70% of usable undersea cables in 2026, yet Washington retains the right to restrict where these networks go and who has access. In 2020, U.S. regulators blocked the Hong Kong segment of the Pacific Light Cable Network—a project backed by Google and Meta—forcing the companies to abandon the direct U.S.-Hong Kong link over Chinese espionage concerns, leaving 13,000 kilometers of already-laid cable abandoned and unused on the ocean floor. Even China's frontier labs, training on domestically-hosted infrastructure, remain dependent on American undersea cables wherever their models touch the global internet—sourcing training data or serving users and running APIs outside China.

The article also emphasizes legal and regulatory tools that amplify U.S. leverage. The U.S. CLOUD Act requires U.S.-headquartered cloud providers to produce data within their possession, custody, or control, even when stored outside the United States. Subject to legal processes, U.S. authorities may obtain not only stored data but also a detailed digital trail of AI activity—including users' prompts, models' responses, who used a system, when and where it was accessed, and technical records revealing behavioral patterns. The article describes such power as exerting 'incredible leverage' on the AI ecosystem.

The article acknowledges limitations to U.S. dominance. Manufacturers in Asia produce semiconductors, AI servers, and other essential components; Nvidia's supply chain runs through TSMC, SK Hynix, and Samsung, alongside system integrators such as Foxconn, Quanta, and Wistron. Last month, Nvidia CEO Jensen Huang announced billions of dollars in new investments and contracts in Taiwan and South Korea—in Taiwan ordering advanced chips and packaging from TSMC along with servers and networking hardware from Quanta Computer and others; in Korea signing billion-dollar deals with memory-chip makers SK Hynix and Samsung, plus new tie-ups with LG, Hyundai Motor Group, and Doosan Robotics. However, the article suggests Taiwan and South Korea are unlikely to weaponize their position and will want to maintain earnings within the U.S. tech stack. China is rushing to build parallel undersea fiber-optic networks along its 'digital silk road' to avoid reliance on American dominance. The article concludes that competing AI ecosystems—a U.S.-led one and a China-led one—may eventually emerge with different standards, infrastructures, and governance, but until that alternative fully materializes, the U.S. will keep a firm grip on the wider AI ecosystem.