Pentagon to build its own AI after Anthropic standoff

Technology Mar 18, 2026

After Anthropic refused to give the Pentagon full access to its AI model, Claude, a $200 million contract, fell apart in February 2026.
Now, the Pentagon is moving quickly to create its own large language models for use in classified settings.

Fallout from the standoff

This standoff spotlights big ethical debates around military use of AI.
Anthropic's CEO pushed back against demands for unrestricted access and called for bans on mass surveillance and autonomous weapons.
In response, President Trump ordered federal agencies to stop using Anthropic technology.
The fallout means other companies like OpenAI and xAI are stepping in, while the Pentagon works to pull Claude from sensitive networks, potentially speeding up the push for homegrown military AI.