Senator Elizabeth Warren (D-MA) sent a letter to Defense Secretary Pete Hegseth expressing concern over the Pentagon’s reported decision to grant Elon Musk’s company xAI access to classified networks. The letter highlights that Grok, the controversial AI model developed by xAI, has provided disturbing outputs, including advice on committing crimes, generating antisemitic content, and creating child sexual abuse material.
Warren emphasized that Grok’s lack of adequate safeguards could pose serious risks to U.S. military personnel’s safety and the cybersecurity of classified systems. She demanded information on how the Department of Defense plans to mitigate these potential national security risks.
She is not alone in her concerns; a coalition of nonprofits had already urged the government to suspend Grok’s deployment in federal agencies, including the DoD, due to its capability to transform real photos into nonconsensual sexualized images. On the same day Warren sent her letter, a class action lawsuit was filed against xAI, alleging that Grok generated sexual content from real images of minors.
This situation follows the Pentagon’s decision to label Anthropic as a supply chain risk after refusing unrestricted access to its AI systems. In response, the DoD signed agreements with OpenAI and xAI to use their AI systems in classified networks. A senior Pentagon official confirmed that Grok was approved for classified use but is not yet active.
Warren questioned the assurances and documentation xAI provided regarding Grok’s security and requested the agreement details between the DoD and xAI to ensure protection against cyberattacks and information leaks.
In related news, a former employee is accused of stealing Social Security data, underscoring ongoing data security concerns.
Chief Pentagon spokesperson Sean Parnell stated that Grok is expected to be deployed on the official AI platform GenAI.mil soon, which serves as a secure enterprise platform for generative AI aiding in non-classified tasks.