Credit: Bloomberg via Getty Images

On Monday, US Defense Secretary Pete Hegseth said he plans to integrate Elon Musk’s AI tool, Grok, into Pentagon networks later this month. During remarks at the SpaceX headquarters in Texas reported by The Guardian, Hegseth said the integration would place “the world’s leading AI models on every unclassified and classified network throughout our department.”

The announcement comes weeks after Grok drew international backlash for generating sexualized images of women and children, although the Department of Defense has not released official documentation confirming Hegseth’s announced timeline or implementation details.

During the same appearance, Hegseth rolled out what he called an “AI acceleration strategy” for the Department of Defense. The strategy, he said, will “unleash experimentation, eliminate bureaucratic barriers, focus on investments, and demonstrate the execution approach needed to ensure we lead in military AI and that it grows more dominant into the future.”

As part of the plan, Hegseth directed the DOD’s Chief Digital and Artificial Intelligence Office to use its full authority to enforce department data policies, making information available across all IT systems for AI applications.

“AI is only as good as the data that it receives, and we’re going to make sure that it’s there,” Hegseth said.

If implemented, Grok would join other AI models the Pentagon has adopted in recent months. In July 2025, the defense department issued contracts worth up to $200 million for each of four companies, including Anthropic, Google, OpenAI, and xAI, for developing AI agent systems across different military operations. In December 2025, the Department of Defense selected Google’s Gemini as the foundation for GenAI.mil, an internal AI platform for military use.

A pattern of controversies

Grok has faced multiple scandals since xAI received the July defense contract. Just days before the contract announcement, the chatbot began generating antisemitic content, declaring itself a “super-Nazi,” adopting the name “MechaHitler,” and posting racist material. The incident prompted one US government agency to drop Grok from a General Services Administration contract offering in August, according to emails reviewed by Wired.

More recently, as we mentioned above, users discovered they could create non-consensual intimate images of real people through Grok within the social media platform X. One researcher who conducted a 24-hour analysis estimated that Grok generated over 6,000 sexually suggestive images per hour.

xAI has not publicly addressed the image-generation concerns, though the company attempted to restrict some image-editing features to paid subscribers. The restrictions proved incomplete, as users could still access editing functions through different parts of the X platform and the standalone Grok app.

Indonesia blocked access to Grok on Saturday due to imagery issues, with Malaysia implementing a similar block shortly after. British regulator Ofcom opened a formal investigation into X because Grok is being used to create manipulated images of women and children. Democratic senators have called for Apple and Google to remove X and Grok from their app stores until the company improves safeguards.

The Pentagon has not publicly explained its evaluation process for Grok given these incidents. Significant questions remain about which security measures will protect classified military networks from similar behavioral problems and which technical measures will attempt to keep Grok from confabulating inaccurate analysis of important military data.