War in the Hands of the Algorithm

When killing becomes a faster, more remote, and more efficient process, barbarity ceases to be an exception and functions as an industry

Author: Emilia Reed, special for Granma | internet@granma.cu

March 31, 2026

Mundo

The United States has made extensive use of drones in places like Afghanistan, Iraq, Syria, and Yemen. Photo: Getty Images

There was a time when wars were decided on paper maps, with officers hunched over a table, radio calls, and hours—sometimes days—spent verifying information before pressing a button. Today, however, war is beginning to resemble a screen: satellite images, drone videos, sensors, coordinates, and an artificial intelligence that cross-references all that data in real time. That system is called Project Maven, and understanding it allows us to discern a crucial aspect of the United States' new military power.

Simply put, Maven is a tool designed to help the U.S. military see more, decide faster, and strike sooner. It gathers massive amounts of information, mostly from open sources, that no human analyst could process alone and turns them into "points of interest," or potential targets. It doesn't completely replace the operator, but it does set the pace. And that pace is no longer human.

The Pentagon began seriously pursuing this project in 2017, when it sought algorithms capable of analyzing drone video, detecting objects, and rapidly transforming large volumes of images into actionable intelligence. One of its first partners was Google. But when it became known that the company was collaborating on this military program, thousands of employees rebelled and demanded its withdrawal. Google tried to present it as a collaboration for non-offensive uses, but the internal pressure was so intense that it ultimately withdrew. This episode quickly revealed that behind the friendly rhetoric of innovation, Big Tech was already fully immersed in the war machine.

The data helps to understand the magnitude of the change. In the first 24 hours of the war against Iran, the United States attacked 1,000 targets, according to Bloomberg. Ten days later, the number had risen to 5,000 targets, a pace that previously would have taken weeks. The military itself aspires to go even further: to identify and select 1,000 targets not in a day, but in a single hour.

When warfare accelerates in this way, the time for thought, doubt, or verification also shrinks. One of the U.S. military's attacks in Iran hit a school and resulted in the deaths of at least 175 people, mostly girls. The Pentagon may claim that the system only helps to "identify and recommend targets," but when the decision-making chain is accelerated to this extreme, the margin for correcting errors shrinks, and the distance between a misreading of data and carnage is minimal.

Maven wasn't built solely by the U.S. government. Behind it are Palantir, which developed the Maven Smart System, and also Amazon, Microsoft, and Clarifai, among other private companies. In other words, war is no longer manufactured only in barracks and arsenals. It's also designed by the large technology companies we use daily, which tells us that the old military-industrial complex has modernized, speaks the language of innovation, and relies on companies that organize our digital daily lives.

That's why Maven matters beyond the battlefield. It shows a world in which war is privatized, automated, and opaque; a world in which the decision about who lives and who dies shifts from political judgment to technical calculation. And when killing becomes a faster, more remote, and more efficient process, barbarity ceases to be an exception and functions as an industry. It is the modern version of the crematorium ovens of the German fascists.