Warren presses Pentagon over decision to grant xAI access to classified networks
Sen. Elizabeth Warren noted that Grok, xAI's controversial chatbot, has created harmful outputs for users and poses a potential national security risk.
TL;DR
- Senator Elizabeth Warren sent a letter to Defense Secretary Pete Hegseth expressing concern over granting xAI's Grok access to classified networks.
- Warren cited Grok's history of generating disturbing content, including advice on committing crimes, antisemitic material, and child sexual abuse material.
- She warned of potential risks to military personnel safety and cybersecurity due to Grok's apparent lack of adequate guardrails.
- A coalition of nonprofits and a class action lawsuit have also raised alarms about Grok's outputs.
- The Pentagon recently labeled Anthropic a supply chain risk and has since signed agreements with OpenAI and xAI for use in classified networks.
- A senior Pentagon official confirmed Grok has been onboarded but is not yet in use in a classified setting.
- Warren requested details on xAI's security assurances and how the DoD will prevent cyberattacks and information leaks.
- Pentagon spokesperson Sean Parnell confirmed plans to deploy Grok on the secure GenAI.mil platform for non-classified tasks soon.