Pentagon Reveals xAI Grok Used for Strikes in Iran

The United States Department of Defense revealed that the military used xAI's Grok Gov Model to launch more than 2,000 munitions at 2,000 distinct targets in Iran within 96 hours during Operation Epic Fury. The disclosure appeared in a federal filing and sworn testimony from Cameron Stanley, the Pentagon's chief digital and artificial intelligence officer, who stated that xAI's computing capacity is as fundamental to modern defense as munitions production.

This admission occurred during a federal court case in Mississippi, where the U.S. Department of Justice defended xAI's Colossus 2 data center against a Clean Air Act lawsuit filed by the NAACP. The government argued that the unpermitted gas turbines powering the facility are vital to national security and that shutting them down would threaten AI innovation supporting military operations. The shift to xAI follows the February termination of contracts with Anthropic after that company refused to allow its Claude model to be used for automated strikes or mass surveillance.

The use of AI in targeting has drawn intense scrutiny following a strike in Minab, Iran, that killed at least 175 people, mostly schoolgirls. While Grok is integrated into Palantir's Maven Smart Systems, some reports suggest the Minab strike may have involved outdated data processed via systems using Anthropic's Claude AI. In response, Senator Kirsten Gillibrand proposed legislation to ban AI use in autonomous and nuclear weapons and to ensure human commanders retain control over lethal decisions.