US Reportedly Dropped '2,000 Munitions at 2,000 Distinct Targets' in Iran Using Elon Musk's Grok AI System

The US military used an artificial intelligence system built by Elon Musk's xAI to help drop 'over 2,000 munitions at 2,000 distinct targets' in Iran within four days, according to a statement filed in a Federal Clean Air Act lawsuit in the United States.

Suspicions that advanced AI tools were being folded quietly into American targeting systems have been building since February, when the US–Iran war escalated and a string of deadly strikes hit civilian sites.

The Wall Street Journal previously reported that Anthropic's AI model Claude was involved in recommending targets, though military officials declined to say whether it had been used in specific attacks, including the bombing of a girls' school.

Now a legal filing connected to a very different controversy has dragged Musk's Grok AI chatbot directly into the frame.

Grok AI's Role in US–Iran Strikes Linked to Elon Musk

The disclosure about Grok AI appears in a sworn statement by Cameron Stanley, identified in the filing as the Pentagon's artificial intelligence chief, who is defending Musk and his company xAI in a Clean Air Act case brought by the NAACP. In that statement, reported by multiple outlets, Stanley says Grok AI was instrumental in coordinating a massive strike campaign in Iran.

According to his account, the chatbot helped enable the release of more than 2,000 munitions against 2,000 separate targets in Iran over a 96‑hour period.

He frames Grok's uninterrupted operation as 'a matter of paramount national security,' arguing that the system sits at the heart of highly sensitive Pentagon work.

Stanley's statement further describes Grok AI as one of four AI models 'currently capable of supporting national security applications,' and one of just three products 'equipped to support mission-critical operations' while maintaining strict confidentiality. That language reads less like a tech brochure and more like a formal attempt to convince a judge that anything threatening Grok's uptime would also threaten US security.

None of this is independently verified beyond the court filing. The Pentagon has not publicly detailed which models were used in which strikes, or released technical documentation to back up the numbers. Until that happens, every claim about the scale and exact role of Grok AI in Iran should be treated with a degree of caution.

Civilian Casualties in Iran and the Reach of Elon Musk's Technology

The months of scrutiny over how the US conducted its campaign in Iran, which has killed thousands of people and hit sites far from obvious military infrastructure.

The war surged in February, and with it, accusations that AI‑driven targeting tools were accelerating decision‑making at a pace that human oversight could not realistically match.

Two incidents stand out. On 28 February, a bombing at the Shajareh Tayyebeh girls' school killed at least 175 people, including more than 100 children. Less than a week later, on 5 March, the Azadi Sport Complex was also hit.

At the time, reporting by Futurism noted that officials would not confirm or deny whether Anthropic's Claude had been involved in determining either target.

Stanley's account now implies a broader ecosystem of AI systems was in play, with Grok AI among the tools feeding into 'mission-critical' operations.

The filing does not describe precisely what the chatbot did. In military jargon, 'supporting national security applications' could mean anything from data analysis and target prioritisation to logistics planning. It is not clear whether Grok generated target lists directly, ranked options prepared by human analysts, or simply processed intelligence that others then used.

Iranian casualties, and especially the attacks on civilian sites, are not addressed in Stanley's testimony. His focus is squarely on keeping Grok running.

Clean Air Lawsuit Pulls National Security Into Elon Musk's Corner

The context for all this is a lawsuit that, on the surface, has nothing to do with Iran. The NAACP has sued xAI over alleged violations of the US Clean Air Act.

The civil rights group accuses the company of using portable combustion turbines to power data centres, generating exhaust fumes that it says are choking nearby communities that are predominantly Black and working class.

Residents have complained that these temporary power units turned their neighbourhoods into an industrial zone without consent, and the NAACP is seeking to force xAI to curb emissions or shut the turbines down.

The case places environmental justice arguments directly against the ambitions of one of the world's most influential tech entrepreneurs.

Stanley's intervention is part of a broader defence strategy aligned with the Trump administration, which has been attempting to shield Musk's AI operations. His statement essentially tells the court that restricting xAI's activities would harm US national security because Grok AI is allegedly so deeply woven into the Pentagon's war‑fighting capacity.

Before this, the US Department of Justice had already suggested it could move to dismiss such 'citizen lawsuits' under certain circumstances, again placing government weight on the side of continued AI development over local environmental concerns.

What remains missing is independent scrutiny. The Clean Air case will test whether a national security argument rooted in a secretive AI war effort is enough to override pollution claims from marginalised communities. It will also determine how far Musk's expanding AI empire can lean on America's military posture when its own conduct is challenged at home.