The U.S. Department of Defense has confirmed it deployed xAI's Grok artificial intelligence chatbot to help coordinate more than 2,000 missile strikes against targets in Iran over a 96-hour period, marking the first explicit government admission of using Elon Musk's AI system for active military operations. This disclosure emerged from a Pentagon filing defending xAI from a pollution lawsuit, raising urgent questions about the role of artificial intelligence in life-and-death military decisions and whether adequate human oversight exists.

What Role Did Grok Play in the Iran Military Operations?

According to Cameron Stanley, the Pentagon's chief digital and artificial intelligence officer, Grok is among only four AI models currently capable of supporting national security applications and one of three equipped to handle mission-critical operations in top secret settings. The chatbot worked alongside other AI systems, including the National Geospatial-Intelligence Agency's Maven Smart System, to identify and process targeting information. Stanley stated in court filings that Grok provided essential support allowing "US forces to deploy over 2,000 munitions to 2,000 distinct targets within 96 hours".

Stanley

The Pentagon emphasized that xAI's Grok Gov Model, a specialized version designed for federal agencies, contains features "found in no other frontier AI model," according to Stanley's sworn statement. However, the specific technical details about how Grok contributed to targeting decisions remain unclear, as the AI systems involved do not explicitly create targets but rather work within broader intelligence frameworks to identify potential points of interest for military decision-makers.

The Pentagon

How Are Concerns About AI in Military Decision-Making Being Addressed?

The disclosure has prompted significant pushback from lawmakers and military analysts concerned about the adequacy of human control in warfare. Several Democrats in Congress are proposing new legislation to establish guardrails around military AI deployment. The most notable proposal comes from Senator Kirsten Gillibrand, whose bill would ensure that human commanders retain control over life-and-death decisions and would ban AI use entirely for nuclear weapons, domestic surveillance, and autonomous weapons systems.

"The most critical decisions affecting our national security and the lives of our service members must always be made by human beings, not unaccountable machines. Right now, the Pentagon is moving toward deploying incredibly powerful AI technology without commonsense guardrails in place, which could have catastrophic consequences that make all of us less safe," Gillibrand stated.

Senator Kirsten Gillibrand, U.S. Senate

Military investigators have raised concerns that AI-driven targeting, combined with human error such as outdated target maps, may have contributed to civilian casualties. U.S. military investigators believe American forces were likely responsible for a strike on an Iranian girls' school in Minab that killed at least 175 people, mostly children, in what analysts describe as the deadliest civilian incident since military operations began in February.

• Congressional Oversight Gaps: Members of Congress have questioned whether faulty intelligence led to strikes killing hundreds of civilians, yet top Pentagon officials declined to investigate whether stricter AI controls could have prevented these deaths.
• AI Model Limitations: While AI systems like Grok and Maven assist in processing targeting data, they do not explicitly create targets but rather present information to human decision-makers, raising questions about how effectively humans can evaluate AI-generated recommendations under time pressure.
• Competing AI Systems: The Pentagon relies on multiple AI products for military operations, and the Department of Defense is also embroiled in a legal dispute with AI company Anthropic after the company refused to guarantee its Claude model would not be used for domestic surveillance or autonomous drones.

Why Is the Trump Administration Defending xAI in a Pollution Lawsuit?

The Pentagon's disclosure about Grok emerged from an unusual legal maneuver in which the Trump administration's Justice Department asked a federal court to dismiss a pollution lawsuit brought by the NAACP against xAI. The NAACP sued xAI in April for operating at least 57 gas-burning turbines without air permits to power its Colossus 2 data center in Memphis, Tennessee, located near homes and schools in Southaven, Mississippi. The lawsuit seeks daily fines of approximately $124,000 for alleged pollution violations and an injunction to halt turbine operations until proper permits are obtained.

In its court filing, the Justice Department argued that the lawsuit threatens national security because the Colossus 2 data center trains and develops AI models critical to both the economy and national defense. Stanley's affidavit emphasized that the data center and others like it are "well positioned" to provide critical surge capacity in the event of armed conflict or other national security emergencies. The government stated it would be "severely" impacted by a court ruling preventing xAI from deploying, refining, and upgrading its systems across the Pentagon.

Environmental law experts have expressed alarm at this unprecedented intervention. Michael Gerrard, an environmental law professor at Columbia Law School, noted that the Justice Department is not disputing the pollution allegations but instead making an unusual claim that citizen lawsuits under the Clean Air Act may be unconstitutional. Erika Kranz, a senior staff attorney at Harvard Law School, observed that this marks the first time the U.S. has intervened in a citizen suit against a private defendant arguing for dismissal on national security grounds.

"It's highly unusual. They're not disputing the allegations in the lawsuit; instead, they're making a very unusual claim that citizen suits are unconstitutional," explained Michael Gerrard.

Michael Gerrard, Environmental Law Professor at Columbia Law School

The methane gas turbines at issue produce nitrogen oxides and other toxic pollutants known to cause asthma attacks, chest pain, and long-term lung damage. Tennessee State Representative Justin J. Pearson, a Democrat who lives a few miles from the data center, called the government's attempt to strike down the lawsuit "unconscionable," stating that "the DOJ seeks to remove any recourse Americans have to protect themselves from harm".

What Are the Broader Implications for AI Accountability and Environmental Protection?

This case could set a significant precedent affecting communities nationwide. Columbia Law School's Gerrard warned that if xAI's approach to powering data centers through mobile turbines is upheld, "that's going to be replicated in many other places, and these mobile turbines are horribly polluting and will have a very negative health effect". The intervention also raises questions about whether national security claims could be used to shield other companies from environmental accountability.

Laura Thoms, director of enforcement for Earthjustice, which represents the NAACP in the case, rejected the national security rationale, stating: "It's a desperate attempt to protect wealthy tech companies from obeying the laws meant to protect people from pollution, turning our communities into sacrifice zones so companies can build and profit from data centers quicker". Thoms previously served as assistant chief for environmental enforcement at the Justice Department until August of the previous year.

The timing of this intervention follows SpaceX's recent initial public offering, which made Elon Musk the world's first trillionaire. The government's defense of xAI occurs as new data centers are rapidly spreading across the United States as part of the Trump administration's push for global AI dominance, with communities increasingly organizing to block new facilities over environmental and health concerns.