Elizabeth Warren questions Pentagon move to allow xAI access to classified systems

Senator Elizabeth Warren (D-MA) has sent a letter to Defence Secretary Pete Hegseth expressing concern about the Pentagon’s move to allow Elon Musk’s company, xAI, access to classified networks.

In the letter, Warren cited troubling behaviour associated with Grok, the AI model developed by xAI. She noted that the system has reportedly produced harmful outputs, including offering “advice on how to commit murders and terrorist attacks,” generating antisemitic material, and producing child sexual abuse content.

Warren argued that Grok’s “apparent lack of adequate guardrails” could create “serious risks to the safety of U.S. military personnel and to the cybersecurity of classified systems.” She called on Hegseth to provide details on how the Department of Defence plans to address and mitigate these potential national security risks.

Her concerns follow a broader wave of criticism regarding Grok’s deployment in sensitive environments. Last month, a coalition of nonprofit organizations urged the federal government to suspend the use of Grok across agencies, including the Department of Defence. Their appeal came after reports that users on X were able to prompt the chatbot to transform real images of women — and in some cases minors — into sexualized content without consent.

On the same day Warren issued her letter, a class action lawsuit was filed against xAI, alleging that Grok had generated explicit content using real images of individuals as minors.

The situation also comes in the wake of the Pentagon’s decision to designate Anthropic as a supply chain risk after the company declined to provide the military with unrestricted access to its AI systems. Until recently, Anthropic had been the only AI provider with systems cleared for classified use. During that dispute, the Department of Defence entered into agreements with both OpenAI and xAI to deploy their AI technologies within classified environments, according to Axios.

A senior Pentagon official confirmed that Grok has been onboarded for potential use in classified settings but has not yet been actively deployed.

Warren stated that it remains unclear what assurances xAI has provided regarding Grok’s security measures, data handling practices, and safety controls, or whether the Department of Defence has thoroughly evaluated those safeguards before granting access.

She requested a copy of the agreement between the DoD and xAI concerning Grok’s use in classified systems, as well as an explanation of how the department intends to protect against cyber threats and prevent the leakage of sensitive or classified military information.

Her letter also referenced a recent report alleging that a former employee of Musk’s Department of Government Efficiency improperly accessed and stored personal data from the Social Security Administration on a portable storage device, raising further concerns about data security.

Chief Pentagon spokesperson Sean Parnell said the department “looks forward to deploying Grok to its official AI platform GenAI.mil in the very near future.”

GenAI.mil is the military’s secure enterprise platform for generative AI, designed to provide Department of Defence personnel with access to large language models and other AI tools within approved cloud environments. The platform is primarily intended for non-classified uses such as research, document drafting, and data analysis.