US Military AI Export Plan: Interagency Coordination Exists but Guardrails Are Weak
The 2025 AI Action Plan and EO 14320 create an American AI Exports Program that mandates multi-agency review (Commerce, State, Defense, Energy, OSTP), but still lacks binding enforcement mechanisms. A January 2026 BIS rule formalized a more flexible export policy for advanced chips — a relaxation from prior presumption-of-denial — which complicates the dual-use risk.
The Trump administration's 2025 AI Action Plan and Executive Order 14320 establish the American AI Exports Program to push full-stack AI systems to middle-power allies as a counter to Chinese influence. The reviewer correctly points out that the EO mandates a consultative process: the Secretary of Commerce must evaluate proposals in consultation with the Secretaries of State, Defense, Energy, and the Director of OSTP, and the EDAG (chaired by the Secretary of State) coordinates financing. So the program does not lack interagency coordination on paper — but the coordination is consultative, not binding. There is no mechanism to overrule a Commerce-determined export license if State or Defense objects, and no public transparency requirement for decisions.
On the chip export front, the reviewer is right that earlier language suggesting a 'partial relaxation in January 2026' was imprecise. The cited Mayer Brown legal update describes the January 15, 2026 BIS final rule as 'formaliz[ing] a more flexible license review policy for transactions involving H200- and MI325X-equivalent chips' — effectively a relaxation from the earlier blanket presumption of denial for China-bound advanced chips. This policy shift creates a strategic tension: the AI Exports Program is designed to promote exports, while the chip rule reduces barriers for the very hardware that enables military AI. Without enforceable interagency guardrails, the risk of dual-use technology leaking to adversaries remains real.
The humanitarian alternative
A more coherent approach would establish binding interagency review standards for military AI exports, modeled on the conventional arms transfer policy, requiring both Commerce and Defense Department sign-offs. Congress should mandate an annual public report on AI export licenses, including aggregate data on rejected applications and end-user monitoring, to prevent technology diversion. Instead of a zero-sum race to dominate allied markets, the U.S. should invest in multilateral AI governance frameworks that set common safety and security standards for military applications, reducing the risk of an AI arms race.
Falsifiable predictions
What this entry claims will happen, and what data would prove it wrong. The Reckoner revisits these against current reality.
- The AI Exports Program will approve fewer than 20 military AI export licenses to middle-power allies in the next 12 months.
- At least one major allied military will publicly express concerns about dependency on U.S. AI systems within 18 months.
Grounded in
- Tech Stack Diplomacy: Policy Implications of the U.S. AI Export ...
- Administration Policies on Advanced AI Chips Codified, with ...
- America's AI Exports Program - Institute for Progress (IFP)
- AI Technology Export Enforcement: 5 Signals Companies Cannot ...
- The New AI Chip Export Policy to China: Strategically Incoherent ...
- AMERICA'S AI ACTION PLAN | The White House
- US: What the New White House AI Action Plan and Executive Order ...
- White House Launches AI Action Plan and Executive Orders to ...
- AI Action Plan - AI.Gov
- What the US AI Action Plan Means for Export Controls and US ...
Original source — excerpted
news Exporting U.S. Military AI Won’t Be Easy"Where middle-power militaries source their AI is a matter of bipartisan concern in Washington. The White House’s “ AI Action Plan ” describes exporting Am..."