The Trump Administration continues to push for rapid onboarding of advanced AI capabilities following its Executive Order on AI issued on June 2, 2026. On June 5, 2026, the Administration followed its Executive Order with National Security Presidential Memorandum-11 (NSPM-11), “Artificial Intelligence in the National Security Enterprise,” providing further directives to defense and intelligence agencies.  While the June 2 Executive Order takes steps toward broad government review and regulation of frontier models, which may have been overcome by U.S. Government actions, NSPM-11 is centered on U.S. Government AI usage. NSPM-11 focuses on maximizing speed and flexibility while preserving security, accountability, and control of advanced AI systems used by U.S. Government national security agencies.  For companies that develop, license, or integrate AI products for U.S. government customers, NSPM-11 presents new opportunities while foreshadowing heightened expectations around reliability, security, multi-vendor access, and support for national security uses.

Baselining AI Requirements for U.S. Government National Security Functions

The Administration wants fast and secure access to the most advanced technologies. NSPM-11 signals a push to close the gap between AI tools available to the public and those available to the U.S. military and the national security workforce.  To this end, the memorandum directs the Secretary of War and Director of National Intelligence to review and update procurement and development processes before October 3, to ensure that the U.S. government can rapidly onboard AI models from multiple vendors.

Four “pillars” will guide this effort, aiming to prioritize both innovation and security.

PILLAR CORE REQUIREMENTS
Adoption Identify opportunities for operational enhancement; eliminate barriers to deployment; cultivate industry partnerships to drive innovation and fielding.
Adaptation Leverage commercial or open-source AI; commercially or internally customize and optimize applications for national security missions.
Assurance Ensure AI systems comply with USG regulations and are “reliable, robust, steerable, and controllable;” safeguard from interference or degradation by malign actors.
Accountability AI must not censor speech, embed ideological bias, or enable unauthorized surveillance; commanders and directors at all echelons will ensure lawful applications and will remain responsible and accountable.

These pillars and the broader framework in NSPM-11 are concerned not just with developing and adopting AI quickly, but also with ensuring government control, continuity, and resilience across vendors and supply chains.  Consistent with that approach, the memorandum favors access to multiple vendors, rather than overreliance on a single provider. It also seeks to ensure that no outside actor can “prevent use of, disable or degrade or materially modify without Federal Government knowledge or approval.”  Security is likely to extend beyond software vulnerabilities to include other concerns about model theft, adversarial misuse, and threats to supply chains, and will likely be shaped by the ongoing Pentagon work on AI cybersecurity mandated by last year’s National Defense Authorization Act.

Notably, NSPM-11 rolls back Biden-era oversight policies for AI tools used by the government.  The Trump Administration appears focused on keeping pace with AI and autonomous technologies from near-peer competitors while still ensuring human oversight and accountability.  In what is likely a response to the ongoing legal dispute with Anthropic, the memorandum calls for agencies to terminate contracts with companies that thwart Executive policies.

A Policy Overhaul with New Institutional Pathways to AI Procurement and Governance

To implement its objectives at the agency level, the memorandum lays out an ambitious plan for developing policies and procedures to build infrastructure and institutional capacity.  Various department and agency heads are tasked with promulgating robust AI guidelines and establishing new organizations, including: an Executive-wide AI governance and reporting policy, updated procurement procedures, an AI “test range” for national security use cases, an AI for National Security Curriculum, an AI National Security Strategic Reserve of non-governmental AI talent, and several interagency and public-private cooperative models to test security and resilience, identify vulnerabilities, and optimize applications.

Ensuring Responsibility and Accountability in the Use of AI-Enabled Weapons Systems

In the closely related (and quickly evolving) field of AI and autonomy in armed conflict, the memorandum provides just 90 days for the Secretary of War to update Department of Defense Directive (DoDD) 3000.09 (“Autonomy in Weapon Systems”) to “ensure the deliberate adoption of AI systems that respect the chain of command and operational authorities.”  First promulgated in 2012 and last updated in January 2023, DoDD 3000.09 sets out the internal processes and responsibilities for the development, acquisition, testing, and use of autonomous weapons systems by DoW, including those with AI capabilities.  Although the existing Directive provides general frameworks for responsible military AI use—for example, by requiring that personnel exercise “appropriate levels of judgment and care” when deploying AI capabilities and that the use of such capabilities “will be consistent” with various ethical and strategic guidelines—it does not specifically address mechanisms for procurement and adoption of AI systems.

It should be noted that 90 days is an extremely accelerated timeline for any DoD Directive, which typically needs to be reviewed and cleared by a number of stakeholders across the Pentagon.  On AI issues that have law of war implications, like DoDD 3000.09, that’s especially true given the multiple offices and Military Departments with equities in this space.  The outcome is sure to be illuminating as to how various actors in the Pentagon are viewing AI and autonomous weapons usage in armed conflict and how they interpret the mandate in NSPM-11 to ensure the adoption of AI systems that “respect the chain of command and operational authorities,” which on its face could be subject to multiple interpretations.  Additionally, AI usage in armed conflict is likely to be addressed in this year’s NDAA, which might make some of the current requirements in DoDD 3000.09 binding law.  All this to say, AI and autonomy in armed conflict is a dynamic and rapidly evolving area of law and policy to watch.

The Way Ahead and Implications for Industry:

For companies interested in selling AI products or related services to the U.S. government, NSPM-11 shows that the national security sector will become a larger and faster moving market, but not an easier one.  Market participants can take a variety of steps to be best positioned in a new development and acquisition environment.

  • Track new procurement processes and vehicles. Similar to its push for more nimble procurement capabilities surrounding drones, the Trump Administration appears eager to overhaul cumbersome procurement bureaucracies and procedures.  Defense tech companies should watch for new acquisition avenues or vehicles that arise from the guidance in NSPM-11.  Agencies are likely to continue moving away from comprehensive requests that favor legacy prime contractors and towards a competitive marketplace with a variety of products.  These new vehicles and approaches provide opportunities, but contractors must navigate these opportunities cognizant of trends to increase supply chain and cybersecurity at the same time (as we have written about here and here).
  • Embrace customization and interoperability. Defense and intelligence use cases for AI far outpace the government’s ability to develop tailored tools.  To maximize government business segments under NSPM-11, AI companies should position their workforces to customize and adapt technologies to government uses, needs, and compliance requirements.  This will mean staying apprised of opportunities for public-private cooperation and using testing environments to hone specifications and build trust.
  • Diversification as risk mitigation. NSPM-11 indicates a wariness of overreliance on large incumbent contractors.  Unlike legacy procurement programs that require years of R&D and intricate supply chains (think: the F-35 program), AI arguably poses fewer barriers to entry for startups and high-growth companies.  Through recent disputes with companies like Anthropic and SpaceX, the government has felt the pain of relying on a single vendor for frontier technological capabilities.  Expect the AI marketplace for defense and intelligence to be accessible but highly competitive.
  • Synchronizing guidance on autonomy and AI in weapons systems. The Secretary of War has until September 3 to issue an update to DOD Directive 3000.09 on Autonomy in Weapon Systems, which will continue to be reviewed annually as AI systems evolve.  In place of current guidance to use military AI capabilities responsibly, the new DOD Directive 3000.09 may implement more concrete requirements for autonomous weapons systems incorporating AI capabilities.  The frontier of AI-enabled autonomous systems represents a massive potential market, but one in which participants will need to track legal and compliance requirements closely and capitalize on opportunities for collaboration with federal customers.

The new Presidential Memorandum puts pressure on national security departments and agencies to adopt advanced AI capabilities more quickly and from multiple vendors, which may benefit companies with mature products that can be deployed without lengthy customization.  At the same time, companies evaluating this market should think not only about whether their products are technically strong, but also whether they can support interoperability, withstand heightened security scrutiny, provide reliable service over time, and align their contractual and operational practices with the government’s interest in continuity and control. To position for success in contracting with the U.S. Government in this evolving landscape of AI contact Fluet’s Government Contracts team.