A retired general’s warning: America can’t fight the AI arms race on tech it doesn’t control

A retired general warns that America's dependence on third-party AI systems like Anthropic creates critical national security vulnerabilities, as the Pentagon cannot fully control or guarantee the security of rented AI infrastructure. The U.S. military's reliance on external AI providers exposes strategic weaknesses in the AI arms race against adversaries like China and Russia.

The Pentagon's relationship breakdown with Anthropic highlights a fundamental asymmetry in modern military capabilities: the world's most advanced defense apparatus depends on AI systems it neither owns nor fully controls. This dynamic creates operational and strategic risks that traditional procurement models don't address. When a private AI company can restrict military access or modify terms unilaterally, it introduces single points of failure into critical national defense infrastructure.

This tension reflects a broader structural problem in AI development. The most capable AI models are concentrated in private companies that operate under commercial and ethical constraints potentially misaligned with military objectives. As AI becomes increasingly central to modern warfare—from autonomous systems to intelligence analysis—the gap between military needs and civilian tech company constraints widens. China and Russia invest heavily in domestic AI infrastructure specifically to avoid this dependency trap.

For investors and developers, this signals accelerating government investment in domestic AI capabilities and potential reshoring of critical AI infrastructure. Defense contractors will likely pivot toward proprietary AI development rather than integration of third-party systems. Venture capital flowing into sovereign AI initiatives and military-grade AI startups should increase significantly.

The coming years will likely see expanded government contracts for domestic AI development, potential export restrictions on advanced AI models, and regulatory frameworks mandating data sovereignty for defense applications. Companies positioned to provide secure, auditable AI systems to government agencies will capture substantial value, while those dependent on unrestricted civilian AI face increased scrutiny and potential restrictions.

• →U.S. military dependence on commercial AI systems creates national security vulnerabilities that adversaries can exploit through supply chain disruption
• →Private AI companies' values and policies may fundamentally conflict with military operational requirements and strategic objectives
• →Government investment in domestic AI infrastructure will likely accelerate as policymakers recognize sovereignty risks in current arrangements
• →Defense contractors focused on proprietary AI development will outperform those relying on third-party integration
• →Geopolitical competition is driving a bifurcation of AI development: sovereign systems for strategic actors versus open platforms for commercial markets