The United States government has disclosed that Grok, the artificial intelligence system developed by billionaire entrepreneur Elon Musk, was employed in military operations targeting Iran, according to a legal filing reviewed by AFP.
The revelation emerged in a June 15 court brief submitted by the US Department of Justice as it defended the gas turbines powering a massive data centre operated by Musk’s AI company, xAI. The facility is currently facing an environmental lawsuit.
In the filing, federal lawyers argued that shutting down the turbines would endanger “American national, economic, and energy security” because the infrastructure supports artificial intelligence capabilities used in military operations.
To support its case, prosecutors presented sworn testimony from Pentagon AI chief Cameron Stanley, who said Grok had already been integrated into Project Maven, the US military’s AI-assisted targeting programme. The system had previously relied on Anthropic’s Claude model.
According to Stanley, Project Maven’s Maven Smart Systems enabled US forces to deploy more than 2,000 munitions against 2,000 separate targets within a 96-hour period during Operation Epic Fury. He credited xAI’s “Grok Gov Model” with significantly improving operational efficiency.
The disclosure comes as the National Association for the Advancement of Colored People (NAACP) pursues legal action against xAI, accusing the company of running dozens of gas turbines without permits in violation of the Clean Air Act. The civil rights group argues that emissions from the turbines disproportionately affect predominantly Black communities.
xAI, however, maintains that the turbines are temporary and mobile, making them exempt from the regulations cited in the lawsuit.
The court filing also revealed that the US government ended its contracts with AI company Anthropic in February after the firm refused to permit its technology to be used for fully automated military strikes or mass domestic surveillance.
Following that decision, the Pentagon expanded cooperation with competitors including Google, OpenAI and xAI as it accelerated the integration of AI into defence operations.
The military’s increasing reliance on AI has sparked opposition within the technology industry. More than 600 Google employees reportedly urged the company not to provide artificial intelligence tools for classified military operations, while critics continue to warn about the broader risks posed by AI-powered warfare.
Despite the transition, the US government acknowledged in March that Anthropic’s Claude model was still being used in operations related to the conflict with Iran.
Musk, a close ally of President Donald Trump, merged xAI with SpaceX in February. SpaceX later completed what was described as the largest initial public offering in history on June 12.
