The drone war in Ukraine remains largely human-operated for now. We’re seeing glimpses of the future, but we’re not there yet.
Ukraine and Russia are experimenting with artificial intelligence-enabled drones, but total autonomy and full-scale deployment remain limited in combat, researchers say.
AI-enabled autonomy in uncrewed systems has the potential to significantly impact how drones are used on the battlefield, reducing the strain on human operators, bypassing electronic warfare and signal jamming, and speeding up the targeting and decision-making process. It can also analyze data and adapt in real time, which is advantageous in combat.
Not quite autonomous and not being used at scale
Kateryna Bondar, a fellow in the Center for Strategic and International Studies Wadhwani AI Center, reported in March that autonomy, a system’s ability to independently operate in complex environments with limited supervision, isn’t “yet present on the battlefield in the war in Ukraine.”
A new report from Institute for the Study of War expert Kateryna Stepanenko explains that “neither Russia nor Ukraine has leveraged AI/ML drones on the battlefield at scale as of early June 2025,” referring to artificial intelligence and machine learning.
The report notes that both sides are “increasingly integrating ML capabilities with some limited AI adaptations into new drone variants on the path to developing fully AI/ML-powered drones.”
Drones that use machine learning, even with some AI, still largely require involvement from an operator. Machine learning could enable the drones to perform pre-programmed tasks, but they lack the autonomy a true AI capability would provide.
Not an AI revolution, at least not yet
Ukrainian forces have observed Russia flying drone swarms that rely on visual terrain navigation — which uses onboard cameras and maps instead of vulnerable GPS — and can autonomously detect and select targets without operator input.
Late last month, Ukraine said it deployed an AI-enabled “mother drone” that can autonomously send first-person view drones to strike targets. And the Security Service of Ukraine reported that the uncrewed systems used in its Operation Spiderweb attack against Russian aircraft at military airbases earlier this month switched to using AI to complete their mission if they lost signal with the operator. The security service said this was done using AI algorithms and manual operator intervention.
Bondar’s report in March on drones said that the attack systems equipped with AI are three to four times more likely to hit their target than drones piloted solely by humans.
But has the revolution come?
One commander of another drone unit said last fall that he expected AI-enabled drones that didn’t need a pilot to be on the battlefield within six months. That future is not here just yet. Earlier this year, a front-line Ukrainian drone unit told Business Insider that AI-enabled drones weren’t being widely used yet.
The war has become a proving ground for cheap drones and emerging technology; however, turning prototypes into a scalable, battlefield-ready AI fleet will require data, chips, and coordination that neither side fully has at the moment.
For Russia, Stepanenko wrote, further development of these important combat capabilities will depend on gathering, storing, and managing battlefield data to train the AI for missions, as well as, critically, sorting out how best to identify enemy drones from friendly ones.
Ukraine has already been working on the latter with situational awareness systems like Delta and Kropyva, which Stepanenko reported are similar to the command and control systems the US Department of Defense has envisioned.
Delta, for example, gives Ukrainian forces across branches and command levels coordinated intelligence from a variety of different systems, including drones, sensors, frontline reconnaissance, and satellites.
In attempting to overcome the broader development challenges, though, Russia struggles with the centralization of drone innovation and production under the government in a way that could hinder advancement. Ukraine, on the other hand, is struggling with resources. Ukraine also faces problems with a lack of government coordination, computing power, and sustainment.
Amid these challenges, Ukraine’s drone developers have nonetheless become a model for the rest of the world. Companies are working closely with front-line forces to meet their needs, effectively creating relatively low-cost systems at scale that push the envelope in new capabilities, such as drone swarm technology.
But in the meantime, more testing and investment are needed by both sides in this war before autonomous, AI-enabled systems really make their mark on the battlefield.
The anticipated changes they could bring, though, would likely overhaul how drone warfare is fought. Ukraine’s Typhoon drone unit told BI that once they became prolific on the battlefield, they’d completely change how operators use drones for reconnaissance and strike missions on enemy personnel, positions, and equipment, as well as against aerial targets.
Read the full article here
