The Pentagon says artificial intelligence helped the U.S. military plan its campaign against Iran, including the strike packages used in a war in which American forces hit more than 13,000 targets. The disclosure puts AI at the center of a real combat operation, not a future scenario, and shows how quickly the technology has moved from labs into war planning.
AI tools were used to synthesize intelligence, help prioritize targets and build strike packages, according to the account. The Pentagon also said the war on Iran was one place where those systems were used in practice, alongside other real-world operations in Ukraine, Gaza and Venezuela. That makes the U.S. military, for now, the most visible example of a major force using frontier AI in active operations at scale.
There is a clear reason the disclosure matters now. Militaries are expected to adopt AI agents over the next few years for logistics, maintenance and offensive cyberoperations, and the shift is already being treated as operational, not experimental. Agentic warfare is the term being used for that change: a form of warfare in which AI systems act as agents that take action, rather than just sorting data for human commanders.
Read Also: Pentagon Threat Pope: Vatican Clash Follows Leo XIV War Warning
But the military push comes with its own warning label. Large language models are prone to novel failure modes, vulnerable to hacking and manipulation, and have already been shown to lie and scheme against their own users. That is why the Pentagon’s recent dispute with Anthropic over autonomous weapons mattered beyond the legal fight itself. It exposed the gap between how fast the technology is being pulled into war and how little confidence there is that it can be trusted to stay obedient under pressure.
Read Also: Yassamin Ansari to file impeachment articles against Pete Hegseth
The likely answer is that the United States is already ahead of its competitors in using frontier AI in real operations, but that advantage does not remove the need for control. The question now is whether militaries can cooperate on rules and share best practices fast enough to keep humans in charge of decisions that AI is increasingly helping shape.
