The Pentagon has officially acknowledged using Elon Musk's Grok AI chatbot to launch more than 2,000 missiles at targets in Iran within a 96-hour period, according to court filings from the Trump administration. This marks the first explicit public admission that the government is deploying xAI's generative AI system for military strikes, intensifying debate over whether humans maintain meaningful control over life-and-death decisions in modern warfare.

The disclosure emerged in a sworn statement by Cameron Stanley, the Pentagon's chief digital and artificial intelligence officer, filed in defense of xAI against a lawsuit alleging the company's data centers illegally pollute Black communities. Stanley argued that Grok's continued operation is "a matter of paramount national security" and that the chatbot is among only four AI models "currently capable of supporting national security applications".

Stanley

What Role Did AI Play in Civilian Casualties?

The revelation comes amid intense scrutiny over civilian deaths resulting from U.S. military operations in Iran. Military investigators believe American forces, aided by AI-driven targeting systems, were likely responsible for a strike on an Iranian girls' school in Minab that killed at least 175 people, mostly children. This represents the deadliest incident for civilian casualties since U.S. and Israeli forces began attacking Iran in February 2026.

Outside analysts have suggested that the Pentagon's AI-driven targeting, combined with human error such as failure to verify whether target maps were current, may have contributed to the bombing. The targets for "Operation Epic Fury" were identified with assistance from the National Geospatial-Intelligence Agency's Maven Smart System, which uses AI to organize data on dashboards to support military decision-making. While these AI products do not explicitly create targets, they work within Maven to identify potential points of interest for military intelligence.

How Is Congress Responding to AI in Military Operations?

Lawmakers are moving to establish guardrails around military AI use. Several Democrats in Congress are proposing legislation to restrict how the Pentagon deploys artificial intelligence, particularly after top officials declined to investigate civilian deaths that might have been prevented by stricter controls on AI systems.

• Human Decision-Making Requirements: A bill from Senator Kirsten Gillibrand 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.
• Transparency and Accountability: The legislation aims to establish what Gillibrand calls "commonsense guardrails" to keep humans in charge and prevent what she describes as "incredibly powerful AI technology" from being deployed without clear rules of engagement.
• Prevention of Catastrophic Consequences: Supporters argue that without these restrictions, the Pentagon risks deploying AI in ways that could have "catastrophic consequences that make all of us less safe," according to the senator's statement.

"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," said Senator Kirsten Gillibrand.

Senator Kirsten Gillibrand, U.S. Senate

Why Is xAI's Colossus Data Center Central to This Dispute?

The Pentagon's defense of xAI extends beyond military performance. The Trump administration is asking a federal judge in Mississippi to dismiss a lawsuit brought by the NAACP, which claims that xAI is violating the Clean Air Act by operating dozens of gas-burning turbines without proper pollution permits at its facilities.

The NAACP alleges that xAI operates at least 57 turbines powering its Colossus 2 data center without the pollution controls required by the Clean Air Act. In court filings, the Pentagon argued that this data center and others are "well positioned" to provide a "critical surge" in energy capacity during an "armed conflict or other exigent circumstances" affecting national security.

"Data centers powering AI products used by the federal government are a long-term strategic tool vital to maintaining our technological advantage against adversaries," stated Cameron Stanley, Pentagon's chief digital and artificial intelligence officer.

Cameron Stanley, Chief Digital and Artificial Intelligence Officer, U.S. Department of Defense

What Happened With Anthropic's Claude AI?

The Pentagon's reliance on multiple AI systems for military operations has created friction with other AI companies. Anthropic, the maker of Claude AI, failed to reach an agreement with the Pentagon after the company determined that the administration would not guarantee against using Claude for domestic surveillance or autonomous drones.

Following the failed negotiations, the Pentagon designated Anthropic as a "supply-chain risk to national security" that could endanger the company's future government contracts. This designation has sparked an ongoing legal battle between the two parties, highlighting broader tensions between AI developers and military applications of their technology.

The Pentagon's explicit acknowledgment of using Grok for military strikes represents a watershed moment in the debate over AI in warfare. As Congress considers new restrictions and companies grapple with military partnerships, the question of whether humans can maintain meaningful control over AI-assisted military decisions remains unresolved.