The US military just hit more than 1,000 targets in Iran within the first day of strikes. That's nearly double the scale of the "shock and awe" campaign that opened the Iraq war two decades ago.
The difference? AI systems that dramatically speed up how targets are identified and prioritized. At the center of this capability is the Maven Smart System, which has quietly become essential to US military operations.
Journalist Katrina Manson's new book, Project Maven: A Marine Colonel, His Team, and the Dawn of AI Warfare, digs into how this all started. Back in 2017, Maven began as an experiment in using computer vision to analyze drone footage. The goal was simple: help humans process the overwhelming amount of video data coming from surveillance drones.
Then things got messy. Google was the initial contractor, and when employees found out their work was being used for military targeting, thousands walked out in protest. The backlash was intense enough that Google eventually dropped the contract.
But Maven didn't die. It evolved and expanded, moving to other contractors and becoming deeply embedded in military operations. What started as a controversial pilot program is now powering targeting decisions at a scale that would have been impossible just a few years ago.
For anyone working with AI tools, this is a reminder of how quickly experimental technology can scale into production systems with real consequences. The targeting AI that caused internal strife at a tech company in 2017 just helped coordinate the largest single-day strike operation in recent US military history.
