White House Orders Faster AI Development for National Security
The Trump administration will accelerate AI development for military and intelligence operations, while prohibiting its use for unlawful surveillance or censoring free speech. President Donald Trump issued a national security memorandum Friday outlining the policy shift.
The order requires Defense Secretary Pete Hegseth to update weapons autonomy guidance within 90 days. The updated directive must ensure AI systems respect the chain of command and that no single vendor becomes a critical point of failure.
The administration also asked leading AI developers to voluntarily submit their most capable models for government cybersecurity testing before public release. This request reflects mounting concerns in Washington about the security risks posed by advanced AI systems.
The Anthropic Dispute
The memorandum follows a high-profile clash between AI lab Anthropic and the Pentagon. In March, the Pentagon formally designated Anthropic a supply-chain risk after the company refused to remove safeguards preventing its Claude AI model from powering autonomous weapons or mass surveillance operations.
The Pentagon argued it should be able to use the technology as needed, provided it complied with US law. Anthropic maintained its position on the restrictions.
The supply-chain designation was an unusual rebuke of an American technology company that had supported military operations, including in Iran.
What the Memorandum Requires
Michael Kratsios, director of the White House Office of Science and Technology Policy, said the order accelerates AI adoption across multiple vendors and updates Defense Department guidance on autonomous weapons to match current technological capabilities. It also prevents any entity from disabling or degrading AI systems that warfighters depend on without prior approval.
Trump said AI technologies should not be developed or used by national security agencies "to censor free speech" or conduct unauthorized or unlawful surveillance.
Trump plans to host a meeting with AI executives as soon as next week to discuss implementation. Anthropic did not immediately respond to requests for comment.
Learn more: AI for Government and Generative AI and LLM applications in the public sector.
Your membership also unlocks:
