Ukraine's Rapid Push to Deploy AI-Enabled Drones for Battlefield Supremacy

24 Jun 2024

The stakes are at an all-time high in the Russo-Ukrainian warfighter’s pursuit of AI-enabled drones for battlefield advantages.

An Oppenheimer Moment For the 21st Century

As The Wall Street Journal recently reported:  “ Electronic warfare is rapidly evolving, pushing drone makers to swap out parts and make craft more autonomous” in Ukraine. Politico wrote that the robot wars have begun and that this is our ‘our Oppenheimer moment.’

As both sides innovate and adapt their drone strategies, the skies over Ukraine and Russia will seemingly be ruled over by autonomous systems.

Ukraine's Minister of Digital Transformation, Mykhailo Fedorov, recently said in an interview that prototypes of AI drones may appear on the battlefield before the year’s end. However, Ukraine has already reportedly used AI-enabled drones to strike Russia, pushing forward in its own pursuit of the upper hand in weaponized AI.

Analysts and technologists are now signaling alarm over the real-time demand for an answer to the AI supremacy question, “reckoning” with the Oppenheimer moment we are brought to once again. As Oppenheimer quoted the Vedas, death and the destroyer of worlds may lurk behind the forced use of AI-enabled force.

Year three of the Russo-Ukrainian forever-war the a moment where that fight is brought to an abrupt forefront of civil thought weighing who will be the victor in the deciding battle of the AI age.

The Fight For Supreme Hive Mind

Evolution in AI-enabled warfighting set siege to modern strategy systems, pushing a game of cat and mouse over which side will reach AI supremacy first, with each side striving to counter the other's aerial assets.

Called the “goldmine” for AI-enabled warfighting tech by Analytics Vidhya as long ago as last year, the stakes for Ukraine’s push to gain AI mastery are higher yet. The impact has reflected on the whole earth, with warfare analysts weighing the odds of what supremacy over the hive mind of an AI-powered war machine means for the world.

Futurists have questioned the end of civilization, wondering if the probability of humanity’s demise now lies in the balance of a weaponized thinking machine. Earlier this year, the Lieber Institute at West Point even called out AI supremacy as “not only dangerous but morally inexcusable.”

In the case of Russia versus Ukraine, a civilization stands or falls on the teeter-totter balance of this “dangerous and inexcusable” concept, as the nature of this new kind of warfare has been tested in the endurance run of Chasiv Yar.

The Stand For Chasiv Yar

Drones, such as First-Person View (FPV) drones have been one of the reasons why Ukraine is still holding on in Chasiv Yar, despite the lack of artillery fire. One Russian soldier recently pleaded on social media for pump shotguns to help stop Ukrainian FPV drones because “they’re simply burning us.”

Earlier this year, Russia had attempted to “sneak through pipes,” a report from Forbes detailed, however, the goal to capture the town ultimately failed. Ukraine “thwarted” the attempt, Euromaidan Press reports, by employing “innovative” drone surveillance. Because of the use of drones, Russia made “no advances” on Chasiv Yar’s vital canal, and the endurance of the stronghold continued.

In the hours that followed the sabotaged waterway attack, Ukraine deployed a volley of missiles and drones that reportedly killed six people inside of Russia. Drone defense of Chasiv Yar flipping to an offensive serves as an example that drones, as units unto themselves, have an evolutionary impact on rapidly changing the status quo of combat.

However, even as technology such as drones can reverse the status quo rapidly, the advantages of this technology can be rapidly force-stopped as observed in Avdiivka.

Advanced Jammers Block Comms

To counter Ukraine’s heavy reliance on drones, Russia has increasingly turned to sophisticated electronic warfare (EW) measures. By deploying advanced jamming systems along the frontline, Russian forces aim to disrupt the communication and navigation systems that Ukrainian drones depend on, particularly GPS signals.

As Russia opened a new front in Kharkiv in May 2024, The Washington Post reported that on the opening morning of the attack, Ukraine’s 125th Territorial Defense Brigade lost all its video feeds due to Russian jamming. The Starlink satellite devices that the Ukrainian military relies on for communications also lost connectivity. Drone feeds from the reconnaissance drones that were used to monitor the entire front disappeared.

These jamming efforts are designed to sever the link between the drone and its pilot, effectively blinding the operator and rendering the drone useless. Electronic warfare targets drones by effectively overwhelming the radio frequencies used to control them. This method disrupts the signals being sent to the UAV, blocking commands from its operator and interfering with the data needed for navigation. By saturating these frequencies with noise or false information, electronic warfare can render drones inoperable, causing them to lose connection, veer off course, or even crash.

Jammers and the Struggle for Avdiivka

In a prior media interview, Dmytro Lysenko, the late drone pilot from the 109th Brigade, spoke of the intense jamming Ukrainian soldiers faced in Avdiivka, often needing to fly many kilometers outside of the intended flight path to avoid Russian jamming. Lexus (callsign) from the special unit “Kondor” of the 1st Presidential Brigade of the National Guard of Ukraine explained that to evade Russia’s powerful electronic warfare systems like the R-330Zh Zhitel, they fly their drones as low to the ground as possible.

However, Lexus believes the real solution is in AI-enabled drones to truly overcome Russian jamming, a theory that has been tested in the real-time back-and-forth of combat realities.

AI Scales Up Conflict Whiplash

As part of Kyiv’s recent drone barrage against Russia’s energy industry, CNN reported that Ukraine was using AI-enabled drones for the strikes. Ukraine has been able to use drones to strike at Russia's Tatarstan region, over 1,300km from Ukraine’s border.

Lexus said that his unit makes AI modifications to their FPV drones that cost between 100 to 200 USD. The AI models are trained by volunteer engineers who gather data and use the footage from the drones to train the models to improve targeting.” Computer vision is used to capture the target and lead the drone to strike directly at it,” said Lexus.

In an interview with The Telegraph, Kateryna Chernohorenko, Ukraine’s deputy defense minister, believes that AI systems “would make one-way attack drones less susceptible to Russian electronic blockers and allow pilots to operate them further from the front line.”

In an interview, Fedir Martynov, the Head of TAF Drones Innovation Hub in Ukraine, said that deploying AI-enabled drones will render Russian electronic warfare helpless in the face of autonomous systems. The race for Ukrainian drone operators now is to quickly train these models to be able to successfully differentiate objects on the battlefield.

“Working on the last mile objection tracking with our AI models in the drones. That is where the market is now,” said Martynov. He believes that one of their partner companies will soon deploy AI-enabled drones that solve the last-mile problem.

“The algorithm isn’t a problem; the biggest problem for us is getting enough training materials for the model,” he said. The faster they can train the models, the faster they can deploy the autonomous drones on the battlefield.

Martynov also highlighted that the models simply require more and more data. Footage of how tanks drive, and how soldiers move around, all this data is needed for the models to improve its targeting. This also gives Ukraine an edge in the Western world in building drones since Ukraine is getting all this data in real-time themselves.

It’s not only battlefield drone models that are hungry for more data, AI models are always hungry for more data, including the world’s most powerful models such as the ChatGPT chatbot. Some companies are creating “synthetic data,” data created by other AI models to train on.

The New York Times highlighted that “The race to lead A.I. has become a desperate hunt for the digital data needed to advance the technology.” The same logic applies to the race to build advanced AI drones on the battlefield.

Martynov also said they’ve received multiple inquiries from companies in the United States. If someone abroad is building drones and wants to weaponize them and validate their solution, it needs to be brought to the battlefield in Ukraine.

Oleksiy Tymofeev, a drone unit commander from the 108th Separate Territorial Defense Brigade fighting is interested in using AI-enabled

drones not only to overcome electronic warfare but to have autonomous drones conduct swarm attacks.

Factoring the Human Question Among Autonomous Systems

However, in a prior interview done with Lexus, he spoke of the growing AI drone arms race on the battlefield. Whoever is first to bring fully autonomous drones to battle will have a clear advantage. But this introduces significant moral questions, Lexus highlighted: "What happens if the drones target civilians? What does Ukraine do if Russian soldiers begin dressing as civilians to avoid targeting?"

Lexus reiterated the confounded sensibility of modern warfare analysts in his observation: "The stakes are monumental. If we lag in pioneering these autonomous drones, adversaries like the Russians might outpace us, and they do not care how many innocent people they kill," the said.

Paul Lushenko, Assistant Professor and Director of Special Operations at the US Army War College, said: “Ukrainian officials better slow down to train these models to distinguish between legitimate and illegitimate targets since it risks losing international legitimacy if it doesn't, which has important implications for the will of its primary patron - the United States - to continue to enable its operations with military hardware and other assistance.”

A Global Problem

Lushenko mentioned the case of Israel and its controversial use of AI targeting in Gaza, with innocent people being caught in the middle. “Their use of AI has resulted in a lot of collateral damage, which has hurt their reputation,” he said.

Lushenko said that fully autonomous drones won’t be a silver bullet for Ukraine either but acknowledged that the rapid development of AI-enabled drones has quickly “outpaced the doctrine” that we have in place. “Whichever side can acquire a target first with AI will have an advantage,” he added.

For Ukraine to overcome Russian jamming, even if using fully autonomous systems, Ukraine's defenders express the need to find those targets faster than the Russians can react to protect them.

However, relying on autonomous systems is expected to introduce new risks. The enemy will seek to break that connection—a physical wire or wireless— to disrupt operations. After that, the next option would be to then opt for more autonomy in the drones. This introduces new vulnerabilities, such as the system being hacked through a cyber-attack and a nightmare scenario where the drone is turned against the original operator.