The Pentagon now has its own version of ChatGPT and Grok
The United States Department of Defense has added official military versions of ChatGPT and Grok to its secure internal portal, GenAI.mil. These tailored tools provide 3 million military and civilian personnel with access to generative artificial intelligence models without exposing sensitive government data to standard consumer data collection. The platform, which previously hosted Google Gemini, has already registered over 1.7 million unique users across the department.
Key Takeaways
- The Pentagon has introduced "ChatGPT Mil" and "Grok for Government" to its centralized GenAI.mil portal, expanding the generative tools available to its 3 million personnel.
Built to prevent sensitive government data from routing through commercial channels, the portal exempts military users from typical consumer data collection practices.
- OpenAI's offering focuses on unclassified routine duties like administrative planning and file management, whereas the Starshield AI version of Grok targets operational tasks including supply-chain management and market research.
The deployment highlights how defense organizations establish isolated environments to leverage commercial frontier models while attempting to mitigate security risks.
- The platform has seen rapid adoption, attracting more than 1.7 million unique users since its launch.
However, the absence of Anthropic from the portal demonstrates the friction between government agencies seeking unrestricted access to technology and developers enforcing safety guardrails.
- The Department of Defense expanded its GenAI.mil platform by adding custom government versions of ChatGPT and Grok alongside Google Gemini.
More than 1.7 million unique users out of 3 million total personnel have already onboarded onto the secure portal.
- Anthropic's Claude model is notably absent after the administration designated the firm a supply-chain risk due to disagreements over safety guardrails.
Stats & Key Facts
- #These tailored tools provide 3 million military and civilian personnel with access to generative artificial intelligence models without exposing sensitive government data to standard consumer data collection.
- #The platform, which previously hosted Google Gemini, has already registered over 1.7 million unique users across the department.
- #The Pentagon has introduced "ChatGPT Mil" and "Grok for Government" to its centralized GenAI.mil portal, expanding the generative tools available to its 3 million personnel.
- #The platform has seen rapid adoption, attracting more than 1.7 million unique users since its launch.
The Pentagon has introduced "ChatGPT Mil" and "Grok for Government" to its centralized GenAI.mil portal, expanding the generative tools available to its 3 million personnel. Built to prevent sensitive government data from routing through commercial channels, the portal exempts military users from typical consumer data collection practices. OpenAI's offering focuses on unclassified routine duties like administrative planning and file management, whereas the Starshield AI version of Grok targets operational tasks including supply-chain management and market research.
The deployment highlights how defense organizations establish isolated environments to leverage commercial frontier models while attempting to mitigate security risks. The platform has seen rapid adoption, attracting more than 1.7 million unique users since its launch. However, the absence of Anthropic from the portal demonstrates the friction between government agencies seeking unrestricted access to technology and developers enforcing safety guardrails.
The Department of Defense expanded its GenAI.mil platform by adding custom government versions of ChatGPT and Grok alongside Google Gemini. More than 1.7 million unique users out of 3 million total personnel have already onboarded onto the secure portal. The military versions exempt defense users from standard commercial data collection practices to protect sensitive government information.
For more details please read the original article at TechCrunch AI.
Continue Learning
Comments
Comments appear only after moderation. Your email identifies your submission to the moderator and is never displayed here.
No approved comments yet.