The Pentagon is shaking up its AI strategy in a major way. After relying heavily on Anthropic’s Claude for classified military systems, the US Defense Department is now opening doors to Elon Musk’s xAI and its chatbot Grok.

Until now, Claude was the only AI model cleared for highly sensitive classified networks used for intelligence analysis, weapons development, and battlefield planning. But tensions grew when Anthropic refused to remove certain usage restrictions, especially around surveillance and autonomous weapons.

Now, Grok has signed an agreement allowing its use for any lawful military purpose inside classified systems. While officials question whether Grok is as mature as Claude, the Pentagon appears determined to diversify AI suppliers.

The decision comes amid rising concerns about operational control, ethical limits, and technological dominance. With AI now embedded in national defense infrastructure, this shift could redefine how the US military balances power, performance, and responsibility in the AI arms race.