Grok AI allegedly used in US Iran strikes, court filing reveals

Elon Musk’s artificial intelligence system Grok has reportedly been used in US military operations involving strikes against Iran, according to a legal filing reviewed by AFP and disclosed on Tuesday.

The filing, submitted on June 15, defends the use of gas turbines powering a large data centre operated by Musk’s AI company xAI, which is currently facing an environmental lawsuit. The case challenges the environmental impact of the facility’s energy infrastructure.

In its legal argument, the US Department of Justice claimed the lawsuit “threatens American national, economic, and energy security” by attempting to disrupt the power supply supporting artificial intelligence systems used in defence-related operations.

Federal prosecutors cited testimony from Pentagon AI official Cameron Stanley, who stated under oath that Grok is already integrated into Project Maven, the US military’s AI-driven targeting program. The system was previously powered by Anthropic’s Claude model.

Read More: Elon Musk’s Grok restricted from creating sexualized AI images

According to the statement, the Maven Smart Systems enabled US forces to deploy more than 2,000 munitions against 2,000 separate targets within 96 hours during what was referred to as “Operation Epic Fury.”

Stanley also highlighted the operational benefits of Musk’s AI technology, praising the “significantly improved efficiency” achieved through the Grok government model.

Meanwhile, civil rights group NAACP has filed a lawsuit against xAI, alleging that the company is operating multiple gas turbines without proper permits, in violation of the Clean Air Act.

The group claims the facilities contribute to pollution in predominantly Black neighbourhoods, while xAI maintains the turbines are temporary and mobile and therefore exempt from regulation.

The controversy comes amid broader tensions over the Pentagon’s growing use of artificial intelligence in military operations. Earlier this year, the US government ended contracts with Anthropic after the company refused to allow its AI systems to be used for fully automated strikes or large-scale surveillance.

Following that decision, the Pentagon reportedly turned to alternative providers, including Google, OpenAI, and xAI, as it expands its AI capabilities in defence applications. However, concerns over ethical use of AI within military systems continue to grow, with internal opposition also emerging within major tech companies.

Reports also indicate that despite policy changes, Anthropic’s Claude model continued to be used in certain military applications as recently as March, reflecting an ongoing transition period in the US defence AI ecosystem.

Elon Musk, a close ally of former President Donald Trump, has reportedly integrated xAI with his aerospace company SpaceX, while the firm recently went public in one of the largest IPOs in history earlier this month.