Ukraine Becomes First To Deploy AI Drone Swarms In Combat

Current attack configurations usually deploy three drones - one for reconnaissance and two for dropping bombs - but swarms of 25 have been tested and swarms as large as 100 drones are expected in the future. 

The recon drone plots a route and the bomber drones then independently determine when to drop. The drones are adaptable, so can adjust if something changes or goes wrong. JL

Ivan Khomenko reports in United24 Media:

Ukraine has become the first military in history to conduct battlefield operations using drone swarms controlled by AI. Ukrainian units have already carried out more than 100 swarm operations. A common formation involves three drones—one reconnaissance platform and two carrying small bombs. The recon drone maps a route while the strike drones decide when and how to release their payload. The technology has been tested with up to 25 up drones. Swarms exceeding 100 UAVs are planned. The Ukrainians are employing software developed by the local company Swarmer, which enables drones to communicate with each other, identify targets, and adapt mid-mission.

Ukraine has become the first military in history to conduct battlefield operations using drone swarms controlled by artificial intelligence, according to The Wall Street Journal on September 2.

The development marks a new phase in the development of unmanned aerial warfare, where groups of UAVs can coordinate and strike with limited human oversight.

Ukrainian forces are employing software developed by the local company Swarmer, which enables drones to communicate with each other, identify targets, and adapt mid-mission.

Analysts cited by the publication say these missions mark the first known routine use of swarm technology in war.

 

The Wall Street Journal (WSJ) reports that Ukrainian units have already carried out more than one hundred swarm operations. A Ukrainian officer confirmed that a common formation involves three drones—one reconnaissance platform and two carrying small bombs.

After being assigned a target area, the reconnaissance drone maps a route while the strike drones decide themselves when and how to release their payload. This approach reduces the number of operators needed: three personnel instead of nine in conventional missions.

Swarmer’s chief executive, Serhii Kupriienko, told the WSJ that the system allows UAVs to adjust if one fails, such as when a battery runs out.

He noted that the technology has been tested with up to 25 drones and that larger-scale trials of swarms exceeding 100 UAVs are planned. The company, partially funded by US investors, initially deployed its software for remote mining before expanding to reconnaissance and precision strikes.

 

The WSJ also notes that the use of artificial intelligence in combat has also raised ethical debates. Current US and NATO policy requires a human decision-maker in the so-called kill chain, but Ukraine’s example shows how AI is increasingly being integrated into real-time battlefield coordination. Swarmer maintains that a human still authorizes the final strike.

Other militaries, including the US, China, France, Russia, and South Korea, are developing swarm technologies, yet Ukraine remains the only country confirmed to be using them regularly in combat.

Experts cited by the WSJ underline that even limited autonomous teaming is a significant milestone, while full-scale swarms of hundreds of drones remain a future objective.

Earlier, General Alexus Grynkewich, NATO’s Supreme Allied Commander Europe, urged defense firms to test their technologies in Ukraine if they want to sell to NATO. He emphasized that Ukraine offers a rare real-world combat environment against a peer adversary, and said feedback from Ukrainian forces is already shaping future procurement strategies across the alliance.