The Sovereign AI Capability Imperative: Why American Dependence on Commercial Frontier AI Constitutes a Structural National Security Vulnerability
The United States leads the world in artificial intelligence, but the leadership rests on four privately-held companies whose foundation models the federal government cannot replicate, cannot fully control, and cannot afford to lose access to without operational degradation. This paper argues that the resulting dependency constitutes a structural national security vulnerability comparable in character to Cold War-era reliance on a small number of strategic suppliers, and that the United States requires a Manhattan-style sovereign artificial intelligence capability — government-directed, government-funded, with government-controlled deployment authority — to address it. The paper situates the argument against the People’s Republic of China’s military-civil fusion doctrine, in which all elements of national capacity are explicitly postured to support military objectives. It steel-mans the principal counterarguments — cost, talent acquisition, government efficiency, the risk of suppressing commercial innovation, and political signaling — and answers each in turn. It concludes with three concrete policy recommendations: a Federally Funded Research and Development Center adapted for frontier artificial intelligence, an Intelligence Community-resident capability for the most sensitive workloads, and a public-private partnership architecture combining government direction with commercial execution under sovereign deployment terms. The recommendations are scoped for implementation within a single administration cycle and draw on the institutional patterns the United States has previously employed for nuclear weapons, satellite reconnaissance, cryptologic capability, and global navigation. The paper is offered as a contribution to active policy discussion in the wake of the 2026 Federal Acquisition Supply Chain Security Act dispute and the seven-vendor classified deployment agreements announced by the Department of War on 1 May 2026.
Paper
Full text
The Sovereign AI Capability Imperative: Why American Dependence on Commercial Frontier AI Constitutes a Structural National Security Vulnerability
Semantic Scholar · 2026
Abstract
The United States leads the world in artificial intelligence, but the leadership rests on four privately-held companies whose foundation models the federal government cannot replicate, cannot fully control, and cannot afford to lose access to without operational degradation. This paper argues that the resulting dependency constitutes a structural national security vulnerability comparable in character to Cold War-era reliance on a small number of strategic suppliers, and that the United States requires a Manhattan-style sovereign artificial intelligence capability — government-directed, government-funded, with government-controlled deployment authority — to address it. The paper situates the argument against the People’s Republic of China’s military-civil fusion doctrine, in which all elements of national capacity are explicitly postured to support military objectives. It steel-mans the principal counterarguments — cost, talent acquisition, government efficiency, the risk of suppressing commercial innovation, and political signaling — and answers each in turn. It concludes with three concrete policy recommendations: a Federally Funded Research and Development Center adapted for frontier artificial intelligence, an Intelligence Community-resident capability for the most sensitive workloads, and a public-private partnership architecture combining government direction with commercial execution under sovereign deployment terms. The recommendations are scoped for implementation within a single administration cycle and draw on the institutional patterns the United States has previously employed for nuclear weapons, satellite reconnaissance, cryptologic capability, and global navigation. The paper is offered as a contribution to active policy discussion in the wake of the 2026 Federal Acquisition Supply Chain Security Act dispute and the seven-vendor classified deployment agreements announced by the Department of War on 1 May 2026.