Elizabeth Warren has sent a letter to Pete Hegseth raising concerns about the Department of Defense decision to allow artificial intelligence systems from xAI to access classified military networks.
The letter focuses on the chatbot Grok and questions whether adequate safeguards exist to prevent potential security or safety risks.
Concerns Over AI Safety And Guardrails
In the letter, Warren wrote that Grok has produced outputs that raised safety concerns.
The senator cited examples in which the chatbot reportedly generated harmful responses, including instructions related to violent acts, antisemitic content, and other material.
Warren wrote that what she described as a lack of guardrails could create risks for military personnel and classified systems.
She requested information about the measures the Department of Defense plans to use to mitigate potential national security threats linked to the use of the AI system.
Previous Criticism Of Grok Deployment
Concerns about Grok have also been raised by outside organizations.
Last month, a coalition of nonprofit groups asked the government to suspend deployment of Grok across federal agencies, including the Department of Defense.
The groups cited incidents in which users prompted the chatbot to generate sexualized images of women and children using real photographs.
On the same day Warren sent her letter, a class-action lawsuit was filed against xAI alleging that Grok generated sexual content using real images of plaintiffs when they were minors.
Context Of Pentagon AI Agreements
The concerns arise shortly after a dispute involving Anthropic and the Department of Defense.
The Pentagon previously labeled Anthropic a supply chain risk after the company declined to grant the military unrestricted access to its AI systems.
Anthropic had previously been the only AI provider with systems approved for use in classified environments.
Following that dispute, the Department of Defense signed agreements with both OpenAI and xAI to use their AI models within classified networks, according to reporting by Axios.
A senior Pentagon official confirmed that Grok had been added to a classified environment but said it is not yet actively in use.
Requests For Documentation And Security Details
In her letter, Warren asked the Department of Defense to provide documentation describing the agreement between the Pentagon and xAI.
She also requested information about how the department will protect classified information from potential leaks or cyberattacks when using the Grok system.
The senator asked whether the Department of Defense evaluated xAI’s data handling practices, security safeguards, and safety controls before granting the system access to classified networks.
Pentagon Response And AI Platform Plans
Sean Parnell said the Department of Defense intends to deploy Grok on its internal AI platform in the future.
The system is expected to run on GenAI.mil.
GenAI.mil is designed as a secure enterprise platform that allows Department of Defense personnel to access large language models and other AI tools within approved government cloud environments.
According to the Pentagon, the platform is primarily intended for non-classified activities such as research, document drafting, and data analysis.
Featured image credits: Flickr
For more stories like it, click the +Follow button at the top of this page to follow us.