Senator Elizabeth Warren has directly challenged Defense Secretary Pete Hegseth over the Pentagon's move to integrate Elon Musk's xAI and its Grok chatbot into classified military networks. In a letter sent Monday, the Massachusetts Democrat cited documented instances where Grok generated advice on violent crimes, antisemitic material, and synthetic child sexual abuse imagery.

Warren's central question is whether the Defense Department has done enough to evaluate the risks. "It is unclear what assurances or documentation xAI has provided... about Grok’s security safeguards," she wrote, warning that inadequate controls could endanger personnel and compromise sensitive systems.

The inquiry arrives amid a significant shift in the Pentagon's AI partnerships. The military recently designated Anthropic a supply chain risk after the company declined to grant unrestricted access to its secure AI systems. Subsequently, the Department of Defense established agreements with both OpenAI and xAI for classified work, according to reports. A senior official confirmed Grok is prepared for classified use but is not yet active on the secure GenAI.mil platform.

Pressure is mounting from multiple fronts. Last month, a coalition of nonprofits called for suspending Grok's federal deployment after users manipulated it to create non-consensual sexual imagery. A class-action lawsuit alleging similar misuse was filed the same day as Warren's letter.

The senator has requested a copy of the DoD's agreement with xAI and a detailed explanation of protocols to prevent data leaks or cyber intrusions. This follows recent, unrelated allegations of data mishandling by a former official in another government department.

Pentagon spokesperson Sean Parnell stated the department anticipates deploying Grok on its official AI platform soon. GenAI.mil is the military's secured system for generative AI, intended for tasks like research and drafting within approved cloud environments.

Source: TechCrunch