Trump Orders Rapid AI Integration Across U.S. Military and Intelligence Agencies
New NSPM-11 restricts vendor control over deployed models while accelerating adoption of advanced AI in national security operations and revising Pentagon autonomy rules
Photo: White House website https://www.whitehouse.gov/gallery/president-donald-j-trump-holds-a-cab…
U.S. President Donald Trump has signed a new National Security Presidential Memorandum (NSPM-11), directing the military and intelligence agencies to accelerate the integration of advanced artificial intelligence models into their operations. The memorandum establishes a framework for the rapid deployment of the most advanced AI models from multiple providers, marking a significant shift in U.S. policy in the field.
According to The Next Web, the most notable provision in the document prohibits companies providing AI models from disabling, degrading, or altering any system relied upon by U.S. security forces without prior approval from the U.S. government. This means a provider would not be able to unilaterally withdraw a model already in operational use, even if concerns arise regarding its deployment.
The directive comes amid a reported dispute between the Pentagon and Anthropic, which was classified as a supply-chain risk after it refused to allow its Claude models to be used for autonomous weapons systems or mass surveillance frameworks.
Michael Kratsios, director of the White House Office of Science and Technology Policy, said that “the men and women who defend our nation deserve the best, most secure, and most reliable artificial intelligence in the world.”
The memorandum tasks Defense Secretary Pete Hegseth with updating, within 90 days, DoD Directive 3000.09, the Pentagon’s core policy governing the use of autonomous and semi-autonomous weapons systems. The existing directive includes requirements for meaningful human judgment before the use of lethal force, and the update is expected to redefine military policy in this area.
Alongside the expansion of AI adoption, the memorandum also sets out general restrictions. It prohibits security agencies from developing or deploying AI models intended to censor free speech, embed ideological bias, or conduct unlawful surveillance against the American people. However, the document does not define these terms in detail or explain how compliance will be enforced.
The memorandum follows a separate executive order signed by Trump earlier in the week, which introduced a voluntary 30-day review window for advanced AI models prior to public release. Together, the two measures signal a dual approach: relatively light-touch regulation of the commercial sector, alongside a rapid acceleration of AI adoption within the national security establishment.
The memorandum also seeks to reduce reliance on single vendors and shift toward a multi-provider model. Responsibility for proper use of AI models will remain with commanders, agency heads, and organizational leaders, who will be required to ensure compliance with policy. The directive also mandates an annual review of key guidelines, in order to keep pace with the rapid evolution of artificial intelligence technologies.