The Pentagon's AI targeting program has quietly achieved a global footprint that would have been unthinkable when Google employees first protested its existence in 2018. Project Maven now operates at more than 130 sites worldwide, with close to 25,000 American military personnel using the platform - a number that more than doubled in the first half of 2025 alone.
According to journalist Katrina Manson's investigation published in WIRED, based on nonpublic documents she reviewed, Maven found its way into 141 exercises and experiments over two years, "Hvylya" reports.
The geographic reach is vast. Maven operates in Australia, Bahrain, Cambodia, Uzbekistan, Vietnam, and Yemen, among others. It maintains two sites in the United Kingdom and analyzes data collected over China, North Korea, Kazakhstan, Myanmar, Pakistan, and Russia. During 2025, Maven had permanent personnel stationed in Japan, Germany, and Qatar, with rotating teams in Jordan, Djibouti, South Korea, and Poland.
A dedicated team supports all six US military services and more than 20 units across seven combatant commands - covering Europe, Africa, the Indo-Pacific, the Middle East, Special Operations, and Space Command. At CENTCOM alone, the region has 13,000 accounts with 2,500 regular users logging in "at least a few times a week," said Rear Admiral Liam Hulin.
Vice Admiral Frank Whitworth, who oversaw Maven's expansion at the NGA, pushed commanders to keep driving adoption. The system's code could be updated in a single day, allowing rapid adaptation to new requirements. "That is the secret sauce of bringing in the operational community in a very, very big way," he said. "Ride that wave."
NGA had accumulated 1 billion AI detections in its computer vision data store by mid-2025. NATO announced plans to adopt Maven Smart System, with ten member nations considering purchases for their own armed forces.
Also read: A British Official's Stark Warning: Ukraine's War Is Coming for "Sleepy, Barely Defended Britain".