Trump Signs NSPM-11 to Accelerate National Security AI and Require Pre-Release Testing
President Donald Trump signed National Security Presidential Memorandum 11 (NSPM-11) on Friday, June 5, 2026, directing U.S. national security agencies to rapidly accelerate AI adoption while revoking Biden-era guardrails (specifically NSM-25) deemed too restrictive.
While the public memorandum focuses on streamlining procurement and enhancing controllability, industry and policy experts agree that the directive is heavily shaped by the administration’s ongoing legal battle with Anthropic.
The Anthropic Feud and Contract Enforcement
In February 2026, Anthropic questioned the reported use of its Claude model by military operators to plan combat operations against Venezuela and Iran. In retaliation, the Trump administration began cancelling Anthropic's federal contracts and transitioning to other vendors, labeling Anthropic a "supply chain risk." Anthropic challenged these cancellations in two parallel lawsuits.
NSPM-11 is designed to prevent such disputes in the future by giving the government a powerful legal cudgel. The memorandum explicitly orders the "termination for default or for convenience" of contracts with any AI developer that attempts to restrict how the military or intelligence community uses its models.1
An exception, however, was carved out for Anthropic’s highly specialized cybersecurity model, Claude Mythos, which is deeply integrated into national security systems.
"But the memo also orders the 'termination for default or for convenience' of contracts with any company that tries to restrict how the government uses its AI2, as the administration alleges Anthropic did... with strictly limited waivers allowed for pressing needs, as reportedly is the case at NSA [with Claude Mythos]." — Breaking Defense
Overhauling DoD Autonomy Directives
A key operational directive in NSPM-11 is a mandate for Defense Secretary Pete Hegseth to submit an updated version of DoD Directive 3000.09 ("Autonomy in Weapons Systems") within 90 days. The directive, last updated under the Biden administration in 2023, governs when and how autonomous systems can make lethal decisions. Tech companies and military planners are closely watching this rewrite to see if the administration maintains rigorous test-and-evaluation safeguards or prioritizes speed at the cost of safety.
"Trump’s memo tasks Defense Secretary Pete Hegseth to submit a new version of 3000.09 within 90 days... 'Speed is great,' [former Project Maven lead Jack] Shanahan added. 'I'm all for speed, but speed without proper test and evaluation is a catastrophe waiting to happen.'" — Breaking Defense
-
An instance of National security now mandates the forced localization of AI development. — Unilateral defense mandates leverage federal purchasing power to force developers into giving the military unrestricted model access. ↩︎
-
An instance of You cannot secure a technology monopoly without blocking foreign human mobility and outbound acquisitions. — The US President's NSPM-11 exploits federal contract terminations to forcefully discipline AI developers that attempt to restrict state military usage of their software. ↩︎