Speed Over Caution: What NSPM-11 Means | Small Wars Journal by Arizona State University

Member Login<br>Become a Member

Search

Advertisement

The Discourse

Speed Over Caution: What NSPM-11 Means

by SWJ Staff

06.08.2026 at 03:18pm

Signed June 5, 2026, NATIONAL SECURITY PRESIDENTIAL MEMORANDUM/NSPM-11 is the Trump Administration’s most comprehensive statement yet on integrating artificial intelligence into U.S. defense and intelligence operations. Rescinding and replacing the Biden-era NSM-25, the memorandum signals a decisive shift in posture from cautious governance first to speed of adoption first.

Structure

The document is organized around four policy pillars and a cascade of timed directives tasking the Secretary of War, the DNI, NSA, OMB, and OPM with reshaping how AI is procured, deployed, secured, and staffed across the national security enterprise.

"Previous administrations imposed undue bureaucracy that hampered the pace of AI adoption, fostered dangerous dependencies on single vendors, and made it challenging for our warfighters to adopt the most advanced technologies."

The memorandum is notable for what it retains from its predecessor (human accountability, civil liberties language) as much as for what it discards (multi-layered pre-deployment review requirements).

We’ve summarized the four pillars for you.

The Four Pillars

Adoption

Eliminate barriers to rapid deployment. Maintain deep industry partnerships. Make frontier models broadly available to national security professionals without delay.

Adaptation

Leverage commercial and open-source AI from diverse suppliers. Customize or internally develop where commercial solutions fall short. Share adapted tools enterprise-wide.

Assurance

Ensure AI is reliable, robust, steerable, and controllable. Contractually prevent vendors from disabling or degrading mission-critical systems without federal approval.

Accountability

AI shall not censor speech, embed ideological bias, or enable unauthorized surveillance. Commanders at every level remain personally accountable for these obligations.

Key Directives — By Timeline

90 days

Update DOD Directive 3000.09 (Autonomy in Weapon Systems) . To be reviewed annually, ensuring autonomous weapons respect the chain of command and operational authorities.

Contract termination authority. SecWar, DNI, and agency heads directed to terminate contracts with companies demonstrating a "repeated pattern of conduct" inconsistent with NSPM-11 policy. Waivers capped at 1 year; must be reported to APST and APNSA.

AI governance policy for national security systems. CNSS and OMB Director to issue policy maximizing consistency with OMB M-25-21 where appropriate.

Classified annex. To be issued addressing sensitive national security dimensions not publicly disclosed.

120 days

Procurement process review. SecWar, DNI, and IC heads to update acquisition to enable rapid multi-vendor onboarding, closing the capability gap between public and national security access to frontier models.

Advanced computing roadmap. Assistant to the President for Science and Technology (APST) and OMB Director to develop a roadmap for compute access, including high-security AI computing facilities and an AI test range for national security use cases.

Private-sector AI security partnerships. SecWar, SecEnergy, DNI, and NSA Director (through AI Security Center) to formalize partnerships for threat intel sharing, joint red-teaming, personnel vetting, and data center security.

Joint AI data and model exchanges. Cross-enclave data/model sharing for common mission applications across the national security enterprise.

AI National Security Strategic Reserve. OPM to establish a reserve of non-governmental AI talent available for federal national security efforts.

AI for National Security Curriculum. DNI and SecWar to develop training ensuring national security personnel understand AI capabilities, limitations, and frontier developments.

Joint AI risk management strategy. DNI, SecWar, and NSA Director to develop baseline AI security practices for the enterprise, submitted for interagency review before publication.

Standardized TEVV methodologies. Test, Evaluation, Verification, and Validation standards for high-security AI systems, to be submitted to APST and the National Cyber Director.

Implications & Takeaways

Acquisition

Multi-vendor mandates reshape procurement. Program managers should expect new contracting vehicles designed for rapid AI onboarding.

Autonomy

DODD 3000.09 revision is the pivotal near-term deliverable. The previous directive’s requirement for appropriate levels of human judgment over the use of force has long generated friction between acquisition speed and legal/ethical review. The 90-day update may relax those thresholds for non-lethal or lower-risk autonomy applications while retaining human-in-the-loop requirements for lethal systems.

Industry

NSA’s AI Security Center is now formally positioned as the center...