US Military Uses AI to Target Over 1,000 Sites in Initial Iran Operation

The US military struck more than 1,000 targets in the first 24 hours of its assault on Iran, nearly double the scale of the shock and awe attack on Iraq over two decades ago. AI systems enabled this acceleration of the targeting process. The Verge reported that one of the thousand targets was a girls’ school, killing more than 150 people, mostly children.

The girls’ school had previously been part of an Iranian naval base, but it was listed online as a school and playgrounds were visible on satellite imagery. The Verge detailed how the Maven Smart System, a chief AI tool used by the US military for targeting, synthesized satellite imagery, radar, social media, and dozens of other data sources to identify and target entities on the battlefield.

Maven speeds up the kill chain by combining computer vision with a workflow management system that finds targets, pairs them with weapons, and allows quick progression through targeting cycle steps. A targeting process that once took hours can now be completed in seconds using Maven.

An official stated that the technology has allowed the US to go from hitting under a hundred targets a day to a thousand targets a day.

With the addition of LLMs, the US can hit up to five thousand targets a day, according to a US official. Project Maven was initiated in 2017 as an experiment in applying computer vision to drone footage. Employee protests at Google occurred in 2018 regarding Project Maven.

Google was the initial contractor for Project Maven and backed out due to the protests. A Google spokesperson said in 2018 that flagging images for review on the drone feed with AI was intended to save lives and was for non-offensive uses only. A Marine intelligence officer pushed forward Project Maven.

The Maven Smart System was built by Palantir. It draws on technologies developed by Microsoft, Amazon, Anthropic, and others. The system is used across the US armed forces. NATO recently purchased the Maven Smart System.

Reporting uncovered military programs to develop fully autonomous weapons, including an explosive-laden drone Jet Ski capable of targeting and destroying targets on its own. The Marine intelligence officer is chief of Project Maven and served as the day-to-day leader.

The officer's motivations for Project Maven stemmed from frustration with poor intelligence tools for US military operators in Afghanistan.

US military operators in Afghanistan fought the war 40 times over every six months due to information not being handed over during troop rotations. Project Maven started in 2017 with funding to use AI against satellite imagery but was repurposed for drone video imagery. At the high point on one day in 2022, the US passed 267 points of interest to Ukraine.

The Verge reported that in supporting Ukraine, US forces used Maven to identify Russian positions, retraining algorithms on new satellite footage to spot tanks in snow after initial failures in desert-adapted models. This marked an inflection point where AI helped speed up artillery operations and targeting.

The US provided points of interest to Ukraine to target Russian equipment and personnel, threading a needle to avoid direct participation by not fully designating targets.

Reporting shows that Maven's integration reduced human roles in the targeting loop from six steps to two, with AI handling assessments and data collection. Central Command stated that AI sped processes from days or hours to seconds. Military ethicists noted risks of gamification in war, where operators might trust on-screen targets without fully understanding supporting data.

Pushback highlighted that the system allows better-tagged data and greater transparency for auditing. The enormous operation in Iran serves as a case study for accountability in using the platform. Debates within the US military weigh leaning into AI acceleration against preserving human assessment to save lives.

Retired Defense Secretary Jim Mattis stated that targeting is no substitute for strategy, emphasizing that hitting many targets does not ensure victory. An example from 1999 involved the US strike on the Chinese Embassy in Belgrade due to an outdated map, illustrating risks if AI speeds targeting without updated data.

The direction of travel points to Maven becoming a program of record by the end of September, with Palantir as prime contractor.

Central Command's commander publicly noted on X that AI proved helpful in operations. Reporting details how the Marine intelligence officer envisioned 'white dots' on maps infused with intelligence, driving Maven's creation to bring analytic tools to frontline operators.