As artificial intelligence rapidly advances beyond everyday applications, its reach is now extending into far more consequential domains, including modern warfare. In a development that underscores this shift, the Pentagon has confirmed that xAI’s Grok, developed by Elon Musk’s company, played a direct role in targeting decisions during the US military campaign against Iran involving more than 2,000 missile strikes.
What Grok Actually Did
Grok was not controlling any weapons on its own. Instead, it was used inside the Pentagon’s intelligence system, which collects and processes data from satellites, sensors, and intercepted communications to help human commanders make decisions.
Grok analysed large amounts of real-time battlefield information, including radar data, messages, satellite images, and past strike records. It used this data to create lists of priority targets and suggest the best order for strikes. Tasks that would normally take analysts days were completed in just a few hours. It also predicted how Iran might respond to each strike, helping commanders plan follow-up actions more quickly.
However, final decisions were always made by human commanders. Grok only provided recommendations, humans carried out the strikes.
Why the Pentagon Chose Grok
The decision to deploy a commercial AI system in a live combat environment, rather than a classified, purpose-built defence system, reflects a broader shift in US military thinking. The Pentagon's Defence Innovation Unit has been accelerating the integration of commercial AI into military operations under the Replicator Initiative and related programmes, on the basis that frontier commercial models now outperform legacy defence systems on many analytical tasks.
Grok's edge, according to defence officials cited in multiple reports, was speed of synthesis across heterogeneous data sources, the ability to pull together signals from dozens of different intelligence feeds simultaneously and produce coherent, ranked targeting recommendations faster than any existing military system could.
The Iran Strikes
The US-Israel military campaign against Iran began on February 28, 2026, following months of escalating tensions over Iran's nuclear programme. The operation involved strikes across Iranian territory targeting nuclear facilities, air defence systems, and military command infrastructure. Iran responded by closing the Strait of Hormuz, triggering the most severe disruption to global energy markets since the 1970s. A preliminary ceasefire agreement was reached on June 15.
The scale of the missile campaign, over 2,000 strikes, is significantly larger than most publicly reported figures during the conflict, and the Pentagon's confirmation of Grok's role represents the first official acknowledgment that a commercial AI product was used at this level of operational involvement in a major military engagement.
Ethical and Strategic Questions
The disclosure has reignited debate over AI’s role in lethal decision-making. Supporters say AI-assisted targeting improves precision and can reduce civilian casualties, while critics warn that using commercial AI in live military operations raises serious accountability and legal concerns.
The development has also drawn attention to Elon Musk’s dual role as head of xAI and SpaceX, placing him at the intersection of AI innovation and US defence operations in an unprecedented way.
Currently, no clear regulatory framework governs the use of commercial AI in military targeting. The Pentagon has said it will release a full review of AI’s role in the Iran operation later this year.
