Musk’s Grok was used to launch missiles at Iran

The Trump administration has reportedly used AI technology from Elon Musk’s company xAI in military operations against Iran, joining other systems that have come under scrutiny due to their role in US-led attacks that resulted in the deaths of hundreds of civilians, including children.

The information is contained in a document from the United States Department of Justicepresented as part of a legal dispute related to the gas turbines that power a gigantic xAI data center, the target of a complaint for alleged environmental violations.

According to , the person responsible for AI at the Pentagon stated that the continued operation of the company’s chatbot constitutes “a matter of national security of utmost importance.”

The document includes the testimony of Cameron Stanleyresponsible for the AI ​​area at the Pentagon, who declares under oath that a tool derived from Grok, the Grok Gov Modelis already being used within the scope of Project Maventhe US military’s AI-assisted targeting program.

According to Stanley, the program allowed the U.S. military to launch more than 2000 ammunition against 2000 different targets in a period of just 96 hours.

The person in charge adds that Maven users process “about two billion tokens per day, the equivalent of up to six million pages of information. According to him, this volume makes xAI’s computational capacity essential for the program to function.

A NAACP (National Association for the Advancement of Colored People) sued xAI, accusing the company of operating dozens of turbines without the necessary licenses, in alleged violation of air quality legislation.

The organization maintains that these installations contribute to pollution in neighborhoods inhabited by black populations. xAI, in turn, argues that the turbines are temporary and mobileso they would not be subject to the same regulatory requirements.

At the end of February, the US government terminated its contracts with Anthropic, after the company refused to authorize the use of its tools in fully automated attacks or mass surveillance programs on US citizens.

Following this decision, the Pentagon would have turned to Google, OpenAI and xAI to continue its bet on AI in military operations. In April, more than 600 employees Google demanded that the company stop providing AI technology to the military for use in classified operations.