Line Crossed? Fully Autonomous Drones Kill Russian Soldiers

A Ukrainian drone manufacturer has confirmed that fully autonomous AI drones killed Russian soldiers in a battlefield test two years ago, according to a report in NewScientist by Matthew Sparkes. 

In the mission, ten quadcopters were sent toward the front line near Bakhmut, with instructions to destroy everything they encountered. Per the reporting, two Russian soldiers were killed. No human in the loop– just do a job. 

If true, this is the most concrete confirmation yet that a machine killed a human being based on a decision made entirely on its own. 

The state of play

This was a one-off test. Ukraine’s own rules currently prohibit fully autonomous targeting at the final stage, and the country’s military maintains it adheres to international humanitarian law. 

But the drone-maker, Alexander Kokhanovskyy, believes those rules should change. He is now building a 64-drone autonomous intercept battery. For this tool, humans are required only at the last step. For now.

The UN has called for a ban on such systems, since no binding international law actually exists to address them. The reality is that most major militaries, including the US, are building toward varying degrees of automation in weapons. 

The difference between “AI-assisted” and “AI-decided” is collapsing.

The reckoning

If a machine selects and kills a target with no human confirmation, who is responsible? The programmer? The commander who launched it? Or the government that permitted the test?

As is increasingly the case, technology is ahead of the law. That gap is getting harder to ignore. 

See the original report here: “Fully autonomous drones have killed human soldiers for the first time.”

While you’re here… 

Check out how Ukrainian drone lessons are rubbing off in yet another conflict: “Guerrillas With FPV Drones: Ukrainian Battlefield Lessons for Kurdish Warfare.”