The US military struck more than 1,000 targets in the first 24 hours of the assault on Iran, nearly doubling the scale of the 'shock and awe' campaign in Iraq two decades ago, according to theverge.com.
This rapid escalation was driven by AI systems designed to accelerate the targeting process, most notably the Maven Smart System. The technology synthesizes satellite imagery, radar, social media, and other data sources to identify battlefield entities and speed up the 'kill chain.'
In her new book,Project Maven: A Marine Colonel, His and the Dawn of AI Warfare, journalist Katrina Manson details how the system evolved from a 2017 experiment in computer vision for drone footage into a cornerstone of modern warfare. While Google was the military's initial contractor, employee protests over the partnership prompted the company to withdraw, leading to the involvement of Palantir and technologies from Microsoft, Amazon, and Anthropic.
The lethal speed of automation
Maven combines computer vision with a workflow management system that pairs targets with weapons. According to an official cited by Manson, the technology has allowed the US to increase its daily strike capacity from under 100 targets to 1,000. With the integration of Large Language Models (LLMs), that number could reach 5,000 targets per day.
The speed of these systems has already resulted in civilian casualties. During the first day of the Iran war, an AI-targeted strike hit a girls' school, killing more than 150 people, mostly children. The facility had previously been a naval base but was listed online as a school with playgrounds visible on satellite imagery.
While some discussion focused on potential 'hallucinations' by AI models like Claude, technology historian Kevin Baker told The Guardian that the issue lies in the acceleration of the system itself. "A chatbot did not kill those children," Baker wrote. "People failed to update a database, and other people built a system fast enough to make that failure lethal."
Project Maven was driven in part by Marine intelligence officer Drew Cukor. Manson reports that Cukor sought to replace inefficient Excel and PowerPoint-based intelligence with an analytic tool that could provide real-time data to frontline operators. His vision included 'white dots' on a map—coordinates infused with elevation and intelligence information.
Though Google initially claimed the technology was for non-offensive uses, Manson’s reporting suggests targeting offensive strikes were a primary goal. One source told Manson, "It’s not like we’re doing it for kicks. The goal of the intel is to take out high-value targets."
As the technology matures, the military is moving toward even greater autonomy. Manson uncovered ongoing programs to develop fully autonomous weapons, including an explosive-laden drone 'Jet Ski' capable of independent target destruction.
