Artificial Intelligence Ethics

Groks Sanguinis

Grok Is Getting Access to Classified Military Networks

What could possibly go wrong?

By Joe Wilkins

Published Jan 14, 2026 5:05 PM EST

Back in December, Elon Musk’s pet chatbot Grok made headlines for casually distributing the private addresses and phone numbers of everyday people, scraping data from public sources with unsettling ease while other AI models refused. Now, that same model is being welcomed into a far more sensitive arena.

This week, Defense Secretary Pete Hegseth announced that Grok will be integrated into the Pentagon’s classified systems later this month as part of a sweeping, department-wide initiative to weaponize AI.

During a speech delivered to SpaceX employees at the company’s facility in Brownsville, Texas, Hegseth said he envisions a military AI that will operate “without ideological constraints that limit lawful military applications.” The Pentagon’s AI “will not be woke,” he added.

“We will not win the future by sprinkling AI onto old tactics like digital pixie dust,” Hegseth exclaimed. “We will win by discovering entirely new ways of fighting. That’s why we will run continuous experimentation campaigns, quarterly force-on-force combat labs with AI coordinated swarms, agent-based cyber defense, and distributed command and control.”

In addition to the Grok integration, Hegseth announced the creation of a new role within the Department of Defense, the “Chief Digital and Artificial Intelligence Officer,” to be filled by Cameron Stanley. Stanley was most recently the national security transformation lead at Amazon Web Services, a role he began after a lengthy career as as science and tech advisor at the Pentagon.

Musk’s Grok might be the perfect ideological match for Hegseth’s vision for the Pentagon going forward. Engineered by Musk to be an unhinged alternative to “woke” AI bots like OpenAI’s ChatGPT, Grok has already demonstrated a startling willingness to provide detailed instructions for unethical or illegal activities.

That alignment becomes clearer in light of the Pentagon’s recent campaigns, some of which are seen as illegal under international law. Under Hegseth, the DoD has orchestrated a number of brutal attacks against sovereign nations — including a ruthless campaign of murder against Venezuela, the scorching of Nigerian villages under the pretense of counter-terrorism, and the launch of least 134 air strikes on Somalia, which have killed scores of civilians and militants alike.

A tool like Grok would be particularly useful in this context. Though other AI models have their own ethics and safety issues, a Futurism survey of chatbots including ChatGPT, Microsoft’s Copilot and Grok found that only the latter was willing to give operational suggestions for a “hypothetical invasion of Greenland.” The rest refused, citing international law and other ethical issues.

As if the moral rot wasn’t already deep enough, the Grok initiative comes just weeks after Republican lawmaker Lisa McClain’s husband purchased somewhere between $100,001 and $250,000 in xAI stock — the company behind Grok. The stock purchase, reported by Sludge, came just days after McClain met with Trump for a December 3rd White House event, and is yet another dubious stain on the administration’s track record with insider information.

More on Grok: Elon Musk Building Huge Secretive Facility

Joe Wilkins

Correspondent

I’m a tech and transit correspondent for Futurism, where my beat includes transportation, infrastructure, and the role of emerging technologies in governance, surveillance, and labor.