The Autonomous Arms Race in the Russia–Ukraine War

The war in Ukraine has become increasingly technological. Take unmanned systems: Kyiv has successfully implemented a drone strategy that combines a drone wall along the immediate front, long-range drones that strike deep inside Russia, AI-enabled mid-range drones that target logistics behind the front, and naval drones that help enforce a blockade at sea. Cheap technology, paired with intelligence operations, is also expanding Ukraine’s global reach.

Along the front, it is a constant race to develop countermeasures. One side deploys a new tactic, and the other races to counter it. After Russia lost access to Starlink earlier this year, it doubled down on mesh network technology, making its drones harder to jam because nearby drones would relay the signal to one another. To counter Russian jet drones, Ukraine has developed faster interceptors, equipped with AI-based terminal guidance systems.

With unmanned ground vehicles, Ukraine is integrating AI-powered target detection systems, along with various sensors, enabling robots to identify enemy soldiers and equipment farther away and more accurately.

To accelerate that effort, Ukraine launched the Defense AI Center ‘A1’ within the Ministry of Defense in March. Led by Danylo Tsvok, the centre aims to help build what he describes as an AI army. ‘In modern warfare, the advantage goes to whoever makes decisions faster,’ Tsvok said.

To that end, Ukraine is attempting to integrate as much battlefield data as possible, not only to identify targets and shorten the kill chain, but also to build comprehensive battlefield awareness. ‘The Vezha module in Delta for real-time battlefield video analysis is also powered by AI,’ said Tsvok. Delta is the Ukrainian military’s battlefield management and situational awareness software. ‘AI detects roughly 70 percent of enemy hardware. The target is detected in about 2.2 seconds.’

There is a reason AI companies such as Palantir are interested in working with Ukraine. ‘Ukraine and Palantir have jointly developed a system for detailed analysis of airstrikes, implemented AI solutions to handle large volumes of intelligence data, and integrated these technologies into the planning of deep strike operations,’ Tsvok added.

Earlier in the war, many manufacturers rushed to deploy AI-enabled targeting systems. Not all proved effective. With so many players competing for attention, trusted benchmarks are needed to verify effective systems.

Yaroslav Azhnyuk, chief executive of battlefield AI-module company The Fourth Law, said the defence industry needed standardised benchmarks, similar to those used for large language models, to measure the real battlefield performance of AI-enabled drones. ‘Building those benchmarks is simply engineering,’ said Azhnyuk. ‘It’s like building a rocket,’ he said. ‘Once you’ve done it, it seems straightforward, but getting there takes years of experimentation.’

And in the next phase of the war, software will continue to play an increasingly important role alongside hardware. Autonomy can be scaled more easily when both the software and hardware are inexpensive and effective. The cheapest version of Fourth Law’s TFL-1 module, is available for only US$150 (A$213).

‘Ukraine has established a rapid cycle for creating and refining technologies,’ said Tsvok. ‘This is important because AI systems must be constantly updated to respond to current challenges on the front lines.’

Supporting this effort, Tsvok said more than 200 Ukrainian companies were developing AI-enabled drones, while more than 70 AI and computer vision systems were already in use on the front line.

Lieutenant General Christian Freuding, chief of the German army, said in June, ‘Data is like ammunition,’ and Kyiv is capitalising on that opportunity by conducting data diplomacy. In the long term, Ukraine aims to build a technological shield centred on autonomous systems, while leveraging partnerships with the West, sharing battlefield data, rapidly iterating on new technologies and improving their effectiveness.

‘We truly believe that AI will become mass technology in the army,’ said Tsvok. ‘One of the cornerstones of AI products should be simplicity, so that every service member can use them easily regardless of their level of digital literacy.’

But Ukraine is not alone in this pursuit. Russia is increasingly adopting autonomous tech. Russian Defence Minister Andrei Belousov said in June that the Kremlin was attempting to advance AI adoption in the army.

According to Ukrainian drone specialist Serhii ‘Flash’ Beskrestnov, new variants of Russia’s Geran Seeker and Molniya strike drones use onboard AI for target recognition and autonomous terminal guidance. Equipped with mesh networking, they are far more resistant to electronic warfare and, Flash says, are already being deployed extensively. Additionally, Moscow is extending the same Seeker architecture to Gerbera drones, while working to advance autonomous capabilities in Shahed drones.

In the near future, Ukrainian military planners envision autonomous robots leading assaults by clearing enemy positions before troops advance. Soldiers would rapidly deploy drones at a moment’s notice to hunt Russian targets. Naval drones would lie in wait at sea, activated remotely before autonomously striking their targets. In long-range strikes, AI would fuse battlefield data, identify targets and calculate optimal flight paths, increasing the success of deep strikes inside Russia.

Both sides are racing to deploy more autonomous systems. For now, Ukraine retains an edge, leveraging partnerships with the West and its fast-paced innovation ecosystem. But the race is only beginning. Whichever side gains the greater advantage in autonomous systems will hold the upper hand in the next phase of the arms race.

David Kirichenko is an associate research fellow at the Henry Jackson Society.

Image of a Hornet one-way attack drone: Thomas Dixon/US Army.

AI contributed no ideas to this article – David Kirichenko.

This article appeared originally at The Strategist (ASPI).