White House to accelerate development of AI for 'war fighting'

The White House announced plans to expedite the development and deployment of artificial intelligence for national security and war-fighting applications, aligning with American values. President Donald Trump stated in a national security memorandum that under his administration, the U.S. will responsibly accelerate AI's use in intelligence and combat domains.

“

Under my Administration, the United States can and will responsibly accelerate the use of AI across intelligence and war fighting domains in line with American values.

Defense Secretary Pete Hegseth has been given 90 days to update directives on the autonomy of weapons systems, ensuring the deliberate adoption of AI systems that respect the chain of command. President Trump emphasized that AI technologies should not be developed or used by the national security apparatus to censor free speech or conduct unauthorized surveillance.

This push for AI in defense occurs amid warnings from AI company Anthropic about the growing risk of humans losing control over AI systems. Anthropic highlighted a trend where companies increasingly delegate AI development to other AI systems. If taken far enough, this could lead to AI systems capable of autonomously designing and developing their own successors, a development with profound implications.

“

These trends have huge implications. AI that can build itself would be a major development in the history of technology, one that could bring enormous good for the world in science, healthcare, and beyond.

Anthropic noted that over 80 percent of the code merged into its system as of May 2026 was authored by its own chatbot, Claude. The company projects that by 2027, Claude could perform tasks that currently take humans weeks. While such advancements could bring enormous benefits to science and healthcare, Anthropic cautioned that full recursive self-improvement might increase the risks of losing human control. The company stressed the growing importance of securing, monitoring, and shaping the behavior of these increasingly autonomous systems.

“

But full recursive self-improvement also might increase the risks of humans losing control over AI systems.

The White House initiative also follows a dispute between Anthropic and the Pentagon. The Pentagon had previously designated Anthropic with a formal supply-chain risk, after the company resisted allowing Claude to power autonomous weapons and mass surveillance. The Pentagon maintained its right to use the technology as needed, provided it complied with U.S. law. Despite Claude's improving code, Anthropic staff noted in late 2025 that Claude-written code was still of lower quality than human-written code.

“

If systems are capable of fully building their own successors, the ways we secure them, monitor them, and shape their behavior all grow much more important.