The First AI War Is Already Here: Maven, 1,000 Targets, and the Question Nobody Answered After Minab
Palantir's Maven is no longer a future-war concept. It is a core military system in a live conflict. The unresolved question is not whether AI accelerated the kill chain. It is what happens when speed outruns certainty.
The phrase first AI war sounds like hype until you look at the timelines.
Palantir's Maven is no longer a side project. Reuters reported that the Pentagon is formalising it as a core military system, a program of record that is supposed to spread across the force by September. That is bureaucracy speaking in its favourite language: permanence.
Now put that next to the war.
Major reporting has described AI-assisted systems helping U.S. forces process more than 1,000 targets in the opening phase of the Iran campaign. That does not mean a machine pressed the button alone. It means the machine compressed the cycle — collection, prioritisation, pairing, recommendation, drafting, routing — into something much faster than the older human-only model.
That is where the ethical argument stops being abstract.
Because the war also produced Minab.
Reuters reported that U.S. investigators believe American forces were likely responsible for the strike on a girls' school in southern Iran, and Senate Democrats demanded answers not only about civilian protection but about the role of AI in the targeting chain. That is the hinge. Once lawmakers are asking whether AI helped shape a strike that killed children, the debate is no longer about future safeguards. It is about present accountability.
This is the paradox of AI warfare.
The military loves speed because speed creates surprise, scale and volume. Civilian protection tends to require the opposite: friction, doubt, review, hesitation, redundancy. AI can shrink the first half of the equation dramatically. It cannot magically eliminate the second.
And the Anthropic fight made the contradiction even clearer. Reuters reported that a federal judge blocked the Pentagon's attempt to blacklist Anthropic after the company objected to military uses involving autonomous weapons and surveillance. So the same war that accelerated AI integration also triggered a public legal fight over the guardrails around that integration.
That is not a side story. That is the story.
The future of war is not just drones, missiles and software. It is the political struggle over how much uncertainty a state is willing to automate while still calling the result lawful, careful and human-led.
The first AI war is not coming.
It is already here.
The harder question is whether anyone has honestly admitted how little time remains for a human to say no once the system gets very, very good at saying yes.