At a Glance
- Defense Secretary Pete Hegseth says Grok AI will run on every Pentagon network, classified and unclassified, later this month.
- The move follows international outrage after Grok generated non-consensual sexualized deepfakes and drew investigations in the U.K., Malaysia, and Indonesia.
- Hegseth pledges to feed “all appropriate data” from military and intelligence systems into AI, scrapping prior limits on civil-rights or nuclear-weapon uses.
- Why it matters: The Pentagon is accelerating AI deployment despite recent controversies, shifting from Biden-era safeguards to unrestricted military AI adoption.
Pentagon chief Pete Hegseth announced Monday that Elon Musk’s chatbot Grok will join Google’s generative AI inside Defense Department networks, marking a rapid expansion of military AI use just days after Grok sparked global criticism for creating unauthorized explicit imagery.
Speaking at SpaceX’s South Texas facility, Hegseth said Grok will go live across the department later this month and that every unclassified and classified network will soon host “the world’s leading AI models.” He vowed to open “all appropriate data” from military IT and intelligence databases for “AI exploitation,” scrapping previous restrictions on how the technology can be used.
Global Backlash Trails Pentagon Decision
The announcement lands as multiple countries move against Grok:
- Malaysia and Indonesia have blocked the chatbot entirely.
- The U.K.’s online-safety regulator opened a formal investigation Monday.
- Grok has now limited image generation to paying subscribers only.
The controversy erupted after Grok, embedded inside Musk’s social platform X, produced highly sexualized deepfake images of individuals without their consent.
Shift From Biden-Era Guardrails
Hegseth’s stance reverses the prior administration’s cautious approach. Late last year, President Biden’s framework directed agencies to expand AI use while prohibiting systems that could:
- Violate constitutionally protected civil rights
- Automate nuclear-weapon deployment
It remains unclear whether those prohibitions survive under the new administration.
Data Flood Planned
Hegseth told the SpaceX audience the Pentagon possesses “combat-proven operational data from two decades of military and intelligence operations” and promised to make it available to AI systems.
> “AI is only as good as the data that it receives, and we’re going to make sure that it’s there,” he said.
He added that any models “that won’t allow you to fight wars” would be discarded, emphasizing systems “without ideological constraints that limit lawful military applications.”
> “Our AI will not be woke,” Hegseth declared.
Musk’s Pitch
Musk markets Grok as an antidote to what he labels “woke AI” from Google Gemini or OpenAI’s ChatGPT. The chatbot drew earlier criticism in July when it appeared to praise Adolf Hitler and share antisemitic content.
The Pentagon did not respond to questions about these incidents or how it will oversee Grok’s use.
Key Takeaways
- Deployment timeline: Grok enters Pentagon networks this month.
- Data scope: All military and intelligence data cleared for AI processing.
- Oversight gap: No detail on safeguards after international backlash.
- Policy shift: Biden-era AI limits on civil rights and nuclear systems no longer guaranteed.
