Abdul-Rahman al-Rawi

A 20-year-old student stepping outside his home in western Iraq never knew he had become part of a new era of warfare — one where algorithms help decide who lives and who dies.

Within seconds of leaving his house in Al-Qaim, near the Syrian border, Abdul-Rahman al-Rawi was killed when a US missile struck a car beside him. His death came during a wave of coordinated American strikes in February 2024 targeting militias across Iraq and Syria.

The attack was hailed in Washington as a technological success. But investigations now suggest the student may be the first acknowledged civilian killed in an AI-assisted airstrike.

A precision strike that wasn’t

The bombardment was part of a broader retaliation after militants killed three US troops in Jordan. In a single night, American forces struck 85 targets believed linked to Iranian-backed militias.

Officials later said machine-learning tools helped identify targets. Those systems are tied to Project Maven, a Pentagon programme that uses AI to analyse satellite imagery and flag potential threats.

Yet civilian harm investigators from Airwars concluded that Abdul-Rahman was mistakenly caught in the strike.

Military uncertainty raises alarm

When asked whether AI assisted the strike that killed the student, United States Central Command said it had “no way of knowing”.

For experts, that response is troubling. If targeting data cannot be traced back to its origin, critics warn, accountability in AI-driven warfare becomes dangerously blurred.

Professor Jessica Dorsey, a specialist in military AI, said the admission raises serious questions about record-keeping and oversight within modern strike systems.

Technology outrunning human judgement

The push to automate parts of the battlefield is accelerating. Advanced militaries increasingly rely on AI to speed up what analysts call the “kill chain” — the process of identifying, verifying and striking targets.

But systems such as Project Maven can struggle in complex environments. In some conditions, accuracy rates can drop dramatically, according to reports cited by analysts.

Researchers warn of “automation bias”, where human operators begin trusting machine recommendations without fully questioning them.

A family left with questions

Months after the strike, the US military acknowledged it was “more likely than not” that civilians had been killed. Abdul-Rahman’s family eventually received a brief letter of condolence.

For his brother Anmar, the message brought little comfort. His father has fallen into depression, and his mother suffered a heart attack after the killing.

The debate over AI warfare may rage in Washington and military labs. In Al-Qaim, it already has a name and a face.

AI, US airstrikes and the dead student caught up in a new age of war

AI is going to be the end of all of us, and not just be drones.