The Unseen Infrastructure: How America’s Grip on AI Remains Firmly in Place

The global AI landscape has been dominated by discussions around model performance and accessibility, with China’s advancements often highlighted as a challenge to US leadership. However, this fixation on models masks a more profound reality – the United States maintains control over the underlying infrastructure that enables AI ecosystems.

At its core, this infrastructure consists of hyperscale data centers, cloud computing infrastructure, and subsea fiber-optic cables. These systems are largely owned by US tech giants, making them vulnerable to American export controls and extraterritorial data extraction laws. Furthermore, a massive commercial-military symbiosis has created extensive ties between these entities and America’s defense and intelligence establishment.

Nvidia’s CUDA software layer plays a crucial role in binding together hardware manufacturers like Broadcom and cloud hyperscalers – Amazon Web Services, Google Cloud, and Microsoft Azure. These companies have large contracts with the Department of Defense to deploy frontier AI tools onto classified military networks. The scale of this commercial-military symbiosis is staggering: ten of the world’s largest companies by market cap are US tech firms, representing a concentration in the trillions of dollars.

Washington is racing to invest heavily in AI research and development, with a $90.7 billion surge in federal AI contracting in 2026 alone. However, this infrastructure extends far beyond mere data centers and cloud services. American tech giants like Meta and Google are building their own privately owned subsea fiber-optic networks, covering vast distances and connecting the world’s major economies.

These pipelines account for 70% of usable undersea cables in 2026, yet Washington retains the right to restrict where these networks go and who has access. China, despite its advancements in AI research, remains dependent on American undersea cables wherever its models touch the global internet – sourcing training data scraped from it or serving users and running APIs outside China.

Beijing is racing to build its own parallel undersea fiber optic networks as part of its digital silk road, a move aimed at avoiding reliance on US infrastructure. However, the US also wields significant power through extraterritorial laws, such as the CLOUD Act. This legislation 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, US authorities may obtain not only stored data but also a detailed digital trail of AI activity – user prompts, model responses, access records, and technical behavioral patterns. This grip on infrastructure gives Washington incredible leverage over the global AI ecosystem.

The limitations of this dominance are not without their challenges. The US’s reliance on a narrow range of hardware manufacturers and cloud providers makes it vulnerable to supply chain disruptions or disruptions to these networks. Moreover, the extraterritorial reach of US laws risks alienating allies and sparking international tensions.

As the world becomes increasingly dependent on AI-driven technologies, it is essential to recognize the unseen infrastructure that underpins this ecosystem – and the implications of US dominance for global stability and governance. The AI landscape may appear to be an open competition between models and nations, but beneath the surface lies a complex web of infrastructure and power dynamics that favors the United States.