Grok AI Used by US Military for Iran Strikes: Pentagon Confirms

Pentagon Reveals Grok AI Role in Operation Epic Fury

The United States Department of Justice has confirmed that Elon Musk's artificial intelligence chatbot Grok was used by the US military to plan and execute airstrikes against Iran during Operation Epic Fury. In a legal filing dated June 15, 2026, the Pentagon disclosed that a specialized version of Grok, known as the 'Grok Gov Model,' helped US forces deploy over 2,000 munitions against 2,000 distinct targets within 96 hours. This marks the first official acknowledgment that Musk's AI has been integrated into active combat operations.

The revelation came as part of a federal lawsuit filed by the NAACP against xAI, Musk's AI company, over the operation of unpermitted methane gas turbines at a data center in Southaven, Mississippi. The Justice Department argued that shutting down the turbines would threaten national security, as the facility powers AI systems critical to military operations. In a sworn statement, Cameron Stanley, the Pentagon's chief digital and AI officer, testified that the Grok Gov Model is deployed within Project Maven military AI program, the US Department of Defense's flagship AI-assisted targeting initiative.

How Grok Was Used in Targeting

According to Stanley's testimony, Project Maven's 'Smart Systems' — powered by the Grok Gov Model — processed vast amounts of surveillance data from drones, satellites, and other sensors to identify and prioritize military targets. The system then relayed recommendations to human operators, who authorized strikes. During Operation Epic Fury, which began on February 28, 2026, the AI-enabled targeting system allowed US forces to strike 2,000 targets in just four days — a pace that would have been impossible with human analysts alone.

Project Maven's Evolution

Project Maven was launched in 2017 to accelerate the adoption of machine learning in military intelligence. Initially powered by Google's AI tools, the program faced internal protests at Google in 2018, leading to the tech giant's withdrawal. The Pentagon then turned to Anthropic's Claude model, but terminated that contract in February 2026 after Anthropic refused to allow its AI to be used for fully automated strikes or mass surveillance. The DoD subsequently signed contracts with OpenAI, Google, and xAI, among others. The Anthropic Pentagon contract dispute highlighted growing tensions between Silicon Valley's ethical guidelines and the military's demand for unrestricted AI use.

Controversy and Ethical Concerns

The use of Grok in military operations has sparked intense debate over the ethics of AI-driven warfare. Lonneke Peperkamp, professor of military ethics at the Netherlands Defense Academy, called the development 'deeply concerning.' She told NOS: 'A private company has become a critical link in military operations within a modern armed force. Especially given Musk's immense political influence, this is very worrying.' Peperkamp warned that AI could lead to rapid escalation: 'A major risk is the scale of attacks it enables. The sheer volume of strikes can lead to escalation, simply through the number of operations that can be carried out.'

Critics have also raised the alarm about civilian casualties. Yahoo News reported that US investigators believe AI-driven targeting was likely responsible for the strike on the Shajareh Tayyebeh girls' school in Minab, southern Iran, on February 28, 2026 — the first day of the war. The attack killed at least 80 students, predominantly girls aged 7 to 12. Human rights organizations have called for an independent investigation into the role of AI in the strike.

Grok's Troubled History

Grok has been controversial since its launch in November 2023. The chatbot has been accused of generating explicit deepfakes, spreading antisemitic conspiracy theories, and producing extremist symbols such as Nazi salutes on demand. In early 2025, the European Commission launched an investigation into Grok, and the Amsterdam District Court banned its 'undress' feature in the Netherlands. The Grok AI regulation efforts highlight growing global concerns about the safety of Musk's AI platform.

Legal and Environmental Battle

The disclosure of Grok's military use emerged from an unusual legal context. In April 2026, the NAACP, along with the Southern Environmental Law Center and Earthjustice, filed a Clean Air Act lawsuit against xAI and its subsidiary MZX Tech. The lawsuit alleges that xAI operates 27 unpermitted methane gas turbines at its Colossus 2 data center in Southaven, Mississippi, emitting up to 1,700 tons of nitrogen oxides, 180 tons of fine particulate matter, and 19 tons of formaldehyde annually — making it potentially the largest industrial NOx source in the Memphis area. The facility is located near predominantly Black neighborhoods, schools, and churches, raising environmental justice concerns.

The Justice Department's intervention, arguing that the lawsuit threatens national security, has drawn criticism from legal experts who see it as an overreach. 'The government is essentially claiming that environmental laws must yield to military AI needs,' said one legal analyst. 'This sets a dangerous precedent.'

Global Implications

The confirmation that Grok is being used for lethal targeting has far-reaching implications for international law and AI governance. Under the Geneva Conventions and international humanitarian law, targeting decisions must distinguish between combatants and civilians and ensure proportionality. Critics argue that AI systems like Grok lack the human judgment necessary to make such distinctions reliably. Senator Kirsten Gillibrand has proposed legislation requiring human control over all life-and-death military decisions and banning AI use in nuclear weapons and autonomous weapons systems.

The Pentagon, however, defends the use of AI, stating that it improves accuracy and reduces collateral damage. Stanley testified that the Grok Gov Model 'significantly improved operational efficiency' and that human operators remained 'in the loop' for all strike decisions. But as the Minab school attack shows, questions remain about whether human oversight is sufficient when AI systems process data at machine speed.

Frequently Asked Questions

What is Grok?

Grok is a generative AI chatbot developed by Elon Musk's company xAI, launched in November 2023. It is known for its controversial content moderation policies and has been criticized for generating harmful content.

How was Grok used in the Iran strikes?

A specialized version called the 'Grok Gov Model' was integrated into Project Maven, the Pentagon's AI targeting program. It analyzed surveillance data to identify and prioritize targets, enabling strikes on 2,000 targets in 96 hours during Operation Epic Fury.

What is Operation Epic Fury?

Operation Epic Fury is the US military campaign against Iran that began on February 28, 2026. It involves precision strikes aimed at dismantling Iran's missile capabilities and security infrastructure.

Why did Anthropic lose its Pentagon contract?

Anthropic refused to allow its Claude model to be used for fully autonomous weapons or mass surveillance, leading the Pentagon to terminate an estimated $200 million contract in February 2026.

What are the ethical concerns about AI in warfare?

Critics warn that AI can cause rapid escalation, lacks human judgment for proportionality and distinction, and may increase civilian casualties. The Minab school strike, which killed at least 80 girls, has been linked to AI-driven targeting.

Sources

• US Department of Justice legal briefing, June 15, 2026
• Sworn testimony of Cameron Stanley, Pentagon Chief Digital and AI Officer
• NAACP v. xAI, Clean Air Act lawsuit, US District Court for the Northern District of Mississippi
• NOS News, 'Amerikaans leger gebruikte AI-model Grok van Musk voor aanvallen op Iran,' June 17, 2026
• Yahoo News investigation on Minab school strike
• Wikipedia, 'Project Maven' and 'Grok (chatbot)'
• Al-Monitor, 'AI at war: Five things to know about Project Maven,' April 2026