Federal agencies flagged Grok's safety before Pentagon deal
Federal officials reportedly raised red flags about xAI's Grok just before the Pentagon approved it for classified military use. A January report from the General Services Administration warned that Grok-4 failed federal safety standards and would pose elevated risks without strict oversight. Why it matters: Despite internal safety warnings, the U.S. government is split as defense leaders fast-track Musk's AI into secretive operations.
Federal officials were sounding the alarm on Elon Musk’s xAI tools just as the military was preparing to roll out the red carpet.
According to The Wall Street Journal, officials across multiple federal agencies have raised serious concerns in recent months over the safety and reliability of xAI's chatbot Grok. Yet this week, the Pentagon went ahead anyway, approving Grok for use inside classified settings.
A 33-page warning from the inside
The most pointed pushback came from the General Services Administration (GSA), which oversees federal technology integration.
A Jan. 15 executive summary of a GSA report, reviewed by the Journal, found that Grok-4 does not meet the safety and alignment expectations required for general federal use and for an experimental federal AI platform.
The full 33-page report detailed public-safety incidents and findings from the GSA’s own hands-on testing. Its takeaway was blunt: even limited use across government would demand strict, layered oversight, without which the tool poses elevated and difficult-to-manage risks.
The Pentagon presses ahead
Despite warnings echoing through several agencies, the Department of Defense decided it could not wait.
The Pentagon's decision this week puts xAI squarely at the center of some of the nation's most sensitive, secretive operations, exposing a sharp divide inside Washington over how fast—and how safely—to adopt generative AI.
Why it matters
This is a classic clash between operational urgency and basic safety protocol.
From the Pentagon's perspective, securing access to cutting-edge models like xAI's is paramount to staying ahead of geopolitical rivals. But bringing a model into classified operations when procurement watchdogs have outright flagged its lack of alignment seems like an enormous gamble.
For enterprise teams and outside developers, the takeaway is clear: do not take government adoption as an automatic seal of safety. If federal evaluators believe Grok requires layered, complex babysitting to avoid major slip-ups, private organizations deploying it for mission-critical workflows need to exercise twice as much caution.
While xAI gains a massive strategic win by breaching national security contracts, the reputational downside is just as steep. If Grok falters in high-stakes environments, the backlash could quickly trigger a wider regulatory clampdown across the entire sector.
Sources (1)
- wsj.com(opens in a new tab)Originalwsj.com