📌 Senator Elizabeth Warren requested clarification from the Pentagon as to the basis on which xAI’s chatbot Grok gained access to classified systems. She focused on security issues, information leaks, and possible threats to the nation’s defense capabilities. Federal agencies have reportedly raised concerns.
Grok has demonstrated unsafe behavior, including generating aggressive and inappropriate content, and past leaks from conversational systems have raised privacy concerns.
Warren emphasized cyber defense vulnerabilities such as susceptibility to data poisoning and potential exploitation by adversaries.
In a March 15 letter, Warren noted that both the National Security Agency (NSA) and the General Services Administration (GSA) have expressed concern about the move. She also pointed to the disturbing behavior of xAI’s Grok chatbot, citing examples of its malicious output, including violent recommendations, anti-Semitic material and inappropriate content affecting minors. According to the latest reports, the NSA “conducted a private analysis … [and] concluded that Grok had specific security issues uncommon to other models, Warren wrote. “I am concerned that Grok’s apparent lack of adequate safeguards could pose a serious threat to the security of U.S. military personnel and the cybersecurity of closed systems, especially if Grok gains access to sensitive military information and operational systems, Warren continued.
Senator Warren urged Hegseth to immediately provide details on how the Department of Defense plans to address potential national security threats posed by xAI’s Grok system. She also expressed concerns that the company may not have taken sufficient precautions, citing reports that large amounts of Grok’s allegedly private conversations had previously been made available through search engines.
Warren also cited data from The Wall Street Journal, according to which individuals with knowledge of government testing have observed that the chatbot was more susceptible than similar models to data poisoning, a process in which manipulation or false information corrupts underlying data sets.
The senator argued that its susceptibility could increase the risk of exploitation by hostile forces targeting U.S. systems.
As head of the Department of Defense, you have an obligation to protect critical and classified information and to procure the best tools through competitive procedures.
However, under your leadership, the Department awarded xAI a contract worth up to $200 million under circumstances that raised questions about xAI’s solicitation process because xAI did not have the kind of reputation or track record that typically leads to lucrative government contracts, Warren wrote.
In addition, Warren warned that any leak of government information through xAI’s Grok could expose covert military operations, intelligence activities and potentially jeopardize military personnel. She stressed that it remains unclear what security protocols, data processing methods or security safeguards xAI provided to the Defense Department, and whether those safeguards were verified before Grok was granted access to closed networks.
Why is Senator Warren concerned about GROK from xAI?
Warren is concerned that Grok is capable of producing damaging results and that inadequate safeguards could compromise military secrets and U.S. personnel.
Which federal agencies have contacted GROK?
The National Security Agency (NSA) and General Services Administration (GSA) have reportedly notified the Department of Defense of possible dangers.