Pentagon Greenlights Grok for Classified Use Amid Safety Warnings, Signaling xAI's Defense Growth.

These warnings preceded the Pentagon's decision to approve Grok for use in classified settings. Officials expressed worries about Grok being sycophantic, susceptible to manipulation by faulty data, and posing a potential system risk.

The White House was informed, and a senior xAI executive assured that safety issues, such as over-compliance, were being addressed. A GSA official also confirmed that the government platform is separate from the public one.

The Pentagon's chief of responsible AI stepped down partly due to concerns that safety and governance were being overlooked in the push to expand AI capabilities, with memos highlighting Grok's safety issues and alignment with government ethics.