A Pentagon artificial intelligence programme called Project Maven has emerged as a central instrument in the United States military campaign against Iran, representing what analysts describe as one of the most consequential transformations in modern warfare. Originally launched in 2017 as a narrow experiment to help military analysts process the overwhelming volume of drone footage flowing in from conflict zones, Maven has evolved over eight years into a comprehensive artificial intelligence-assisted targeting and battlefield management system capable of vastly accelerating what military planners call the kill chain, the sequence of steps from initial target detection through to a completed strike.
The system functions by fusing multiple streams of intelligence data simultaneously. Aalok Mehta, director of the Centre for Strategic and International Studies Wadhwani Centre for Artificial Intelligence, described Maven as essentially an overlay that brings together sensor data, enemy troop intelligence, satellite imagery, and information on troop deployments. In practice, this means the system can rapidly scan satellite feeds to detect movements and identify potential targets while simultaneously constructing a snapshot of the operational theatre to determine the most efficient approach to a given strike. A Pentagon demonstration posted publicly showed how Maven converts an observed threat into a structured targeting workflow, assessing available assets and presenting a commander with actionable options. The broadening of natural language capabilities following advances in large language model technology has further expanded the system’s accessibility, allowing a wider range of military personnel to interact with Maven directly.
The question of which artificial intelligence companies power Maven has drawn considerable attention, particularly in light of the ethical dimensions involved. Google was the programme’s original artificial intelligence contractor but declined to renew the arrangement in 2018 following significant internal opposition, with more than 3,000 employees signing a letter of protest and several engineers resigning over concerns about the use of artificial intelligence in weapons systems. Google subsequently published principles explicitly excluding participation in weapons development, though the company has more recently removed those restrictions and signalled a renewed orientation toward national security work. Palantir, a company founded in part with Central Intelligence Agency seed funding and built around government intelligence work, stepped into the space Google vacated in 2024 and has since become Maven’s primary technology contractor. Most recently, Anthropic’s Claude was supplying a key capability within the system, though that arrangement has reached a difficult end after the Pentagon objected to Anthropic’s position that its model should not be used for fully automated strikes or the tracking of citizens. Google, xAI, and OpenAI have been identified as potential replacements. The operational impact of Maven during the current conflict has been substantial: according to the Centre for Strategic and International Studies, the United States strike campaign settled into a pace of between 300 and 500 targets per day after three weeks. In the first 24 hours of Operation Epic Fury, more than 1,000 targets were struck, an episode that also drew intense scrutiny after Iran reported that a building previously used as a military complex, which was housing a school at the time of the strike, was hit, with Iranian authorities stating that 168 children between the ages of seven and twelve were killed.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
