Pentagon: xAI's Grok Model Used to Support Military Operations Against Iran
According to documents filed in a court proceeding in the United States, the Pentagon considers Grok, the generative model developed by xAI, a technology of strategic importance for national security and has utilized it to support military operations conducted against Iran.
The information emerges from a sworn statement signed by Cameron Stanley, the digital and artificial intelligence officer of the Department of Defense. In the document, submitted as part of a lawsuit involving xAI and its computing infrastructure, Stanley states that Grok falls within a select group of AI models currently deemed suitable to support national security applications and classified operations.
According to the documentation, the system has been used in activities that led to the deployment of over 2,000 munitions against as many targets within a span of 96 hours. The documents do not describe the level of autonomy of the model nor do they suggest that Grok made operational decisions independently, but they highlight its role as a support tool within a broader decision-making chain.
The U.S. administration specifies that various AI-based systems are employed to analyze large amounts of data and assist military personnel in identifying elements of interest. Among these is the Maven Smart System of the National Geospatial-Intelligence Agency, a platform that aggregates and presents geospatial information to analysts and operation managers. Such tools do not directly generate military targets but contribute to evaluation and planning processes.
The use of artificial intelligence in warfare is receiving increasing attention from independent observers, human rights organizations, and some members of Congress. In particular, several analyses have hypothesized that automated decision-support systems may have contributed, along with human errors and outdated information, to episodes that resulted in civilian casualties during recent military operations in the region.
The context in which this information emerges is related to an entirely different controversy. The Department of Justice is defending xAI against a lawsuit filed by the NAACP, claiming that the company’s data centers operate utilizing dozens of gas-powered turbines without the environmental permits required by the Clean Air Act. According to the allegation, these facilities contribute to the pollution of surrounding communities.
In its defense, the government argues that xAI's infrastructures represent a strategic component for maintaining the computing capacity needed for the AI applications utilized by federal agencies. Stanley further states that the data centers could provide additional energy capacity in the event of armed conflicts or other emergencies involving national security.
As is known, the Pentagon is also involved in another dispute regarding the adoption of AI technologies, this one with Anthropic, which is still under ongoing litigation.