US says Musk’s Grok AI was used in strikes against Iran

Elon Musk’s Grok artificial intelligence model was used by the United States military during strikes against Iran, according to a government legal filing reported by AFP. 

The disclosure was made in a June 15 brief filed by the US Department of Justice in defence of gas turbines powering a large data centre operated by Musk’s artificial intelligence company, xAI.

The turbines are at the centre of a lawsuit filed by the National Association for the Advancement of Colored People (NAACP), which alleges that xAI operated dozens of turbines without permits and violated the Clean Air Act.

The civil rights organisation says emissions from the turbines are affecting predominantly Black neighbourhoods. xAI maintains that the turbines are temporary and mobile and are therefore not covered by the regulations cited in the lawsuit.

In its filing, the Department of Justice argued that efforts to halt the turbines could affect US national, economic and energy security by cutting power to artificial intelligence systems supporting military operations.

Federal prosecutors submitted sworn testimony from Pentagon artificial intelligence chief Cameron Stanley, who said Grok was being used under Project Maven, the US military’s artificial intelligence-assisted targeting programme.

According to Stanley, Project Maven’s Smart Systems helped US forces deploy more than 2,000 munitions against 2,000 separate targets within 96 hours during Operation Epic Fury.

He also pointed to improved operational efficiency resulting from the use of the Grok Gov Model.

Project Maven initially relied on Anthropic’s Claude artificial intelligence model. However, the US government terminated its contracts with Anthropic at the end of February after the company refused to permit its technology to be used for fully automated strikes or mass surveillance of Americans.

The Pentagon subsequently turned to other artificial intelligence companies, including Google, OpenAI and xAI, as it expanded the military use of the technology.

The transition has not been immediate. The US government acknowledged in March that Claude was still being used in operations related to the war in Iran.

The military use of artificial intelligence has also faced opposition from technology workers. More than 600 Google employees called on the company not to provide artificial intelligence tools for classified military operations.

The latest disclosure emerged as xAI seeks to defend the power infrastructure supporting its data centre against the environmental challenge.

Musk, a close ally of President Donald Trump, merged xAI with his space exploration company SpaceX in February. SpaceX carried out what was described as the largest initial public offering in history on June 12.