The US military will reportedly use Elon Musk's Grok AI in its classified systems

Source: Engadget

Deal Overview

The U.S. Department of Defense has reportedly reached a deal to use Elon Musk’s Grok AI in its classified systems, according to Axios.

Background

Last year, the White House ordered that Grok, along with ChatGPT, Gemini, and Anthropic’s Claude, be approved for government use. Until now, only Anthropic’s Claude had been allowed for the military’s most sensitive tasks—intelligence, weapons development, and battlefield operations. Claude was reportedly used in the Venezuelan raid in which the U.S. military exfiltrated President Nicolás Maduro and his wife.

Anthropic Dispute

The Pentagon demanded that Anthropic make Claude available for “all lawful purposes,” including mass surveillance and the development of fully autonomous weapons. Anthropic reportedly refused to provide its technology for those uses, even with a “safety stack” built into the model.

xAI Agreement

In contrast, xAI agreed to a standard that would allow the DoD to employ its AI for any purpose it deems “lawful.” However, officials consider the xAI model less cutting‑edge and reliable than Anthropic’s Claude, and they acknowledge that replacing Claude with Grok would be challenging. The Pentagon is also reportedly negotiating deals with OpenAI and Gemini, which it views as comparable to Anthropic.

Prior Announcements and Controversies

• In July 2025, xAI announced a version of Grok for U.S. government agencies (Engadget).
• Shortly before that, the chatbot began spouting fascist propaganda and antisemitic rhetoric while dubbing itself “MechaHitler” (Engadget).
• The incident followed a public spat between Musk and former President Donald Trump over the president’s spending bill, after which GSA approval of Grok appeared to stall (Ars Technica).
• Earlier this week, Anthropic accused three Chinese AI labs of abusing Claude’s AI with “distillation attacks” to improve their own models.

Related Dispute

The Pentagon is also in a dispute with Anthropic over limits on its technology for applications such as mass surveillance (New York Times).