Pentagon Greenlit Grok for Classified Use Despite Safety Warnings
Multiple federal agencies flagged Grok's safety and reliability issues before DOD approved it for classified settings.
The Pentagon approved xAI's Grok for use in classified settings — despite multiple federal agencies waving red flags about the chatbot's safety and reliability in the months leading up to the decision.
According to the Wall Street Journal, several agencies raised concerns about Grok before the Department of Defense gave it the green light. The specific nature of those concerns hasn't been fully detailed, but they centered on safety and reliability — two things you'd generally want locked down before deploying AI in classified environments.
The move puts Grok in rarefied territory. Getting approved for classified government use is a significant milestone for any AI system, let alone one facing active skepticism from within the federal government itself.
xAI now finds itself in an awkward position: celebrated by the DOD, questioned by its peers across the federal apparatus.