6 May 2025, NIICE Commentary 10802
Bibek Dhoj Thapa
While everyone was concerned with how the US and the West were arming Ukraine to the teeth with massive firepower, private AI companies were supplying Ukraine with their technology. Palantir Technologies stood out from the rest. Named after J.R.R. Tolkien’s infamous crystal balls, “Palantir” literally translates to palan “far” and tir “watch over” in High Elvish. This company gave Ukraine access to its tech, like the MetaConstellation, and others absolutely free of charge.
Companies like these claim that they were seeing great success while providing their services to various security agencies in the US, but to really test their limits, they needed something massive. This is exactly what the Russia-Ukraine war presented. It was a win-win for both sides, as these companies wanted to test the limits of their technology, and get it ‘battle-tested’, whereas Ukraine was ready to adopt anything that they felt would help them gain the edge over their adversaries.
Ukraine’s Use of AI in Combat Roles
This war might be the first time that AI was used on such a massive scale. MetaConstellation is being used to gather raw intelligence from various sources like drones, satellites, radar, and reports from Ukrainians on the ground and also to analyse images that see through clouds and detect heat signatures to track troop movements and artillery fire. Using AI, it processes this data to help military officials identify the best possible targets and enemy positions.
According to a report by CSIS, Ukraine's primary objective in using AI is to minimise the number of human personnel on the ground. Their use of drones has evolved significantly since the early days of the war. Previously, humans were directly involved in tasks such as analysing drone footage, recognising and tracking targets, navigating drones, and extracting intelligence from sound and text. Now, AI has taken over 99% of these activities, not just to replace boots on the ground but also because of its surprising accuracy.
Ukraine has felt that AI-powered automatic target recognition (ATR) has dramatically improved target identification, extending recognition ranges from 300 meters to an average of 1 km in combat situations and up to 2 km in optimal conditions. It can even detect hidden or camouflaged targets. Additionally, drones equipped with autonomous navigation have significantly increased target engagement success rates from approximately 10–20% to an astonishing 70–80%. This is not the scary part; the fact that these AI models continuously improve themselves after each use at an astonishing rate is terrifying.
Ukraine’s Use of AI in Combat Support and Non-Combat Roles
Along the course of this war, it is seen that current AI technology is better suited for combat support roles rather than for direct engagement. Ukraine’s use of Clearview AI software helped identify around 230,000 Russian personnel involved in the invasion by connecting them to evidence of alleged war crimes. Clearview AI analysed vast amounts of data, linking accusations to satellite imagery, troop movements, and open-source materials like social media photos and videos.
Additionally, Ukraine is using Palantir’s AIP to analyse the status of landmines across its territory. Their focus is on prioritising land with high economic value, reducing hospital travel times, and restoring critical infrastructure such as power, bridges, and schools based on real-time data to benefit the most people. Ukraine is also experimenting with AI and precise satellite imagery to speed up land clearance where there’s no sign of mines. Furthermore, they are using AI to improve evacuation strategies by monitoring schools impacted by air-raid warnings or power outages, tracking road conditions, and estimating how long it would take students to reach shelters equipped with Wi-Fi.
Implications for the Future of Warfare
Ukraine’s use of AI in the war has provided a glimpse of what the future of warfare is going to look like. Firstly, the notion that soldiers are lean and physically active might be replaced, as it is seen that any teenager with the necessary skills and knowledge can wreak havoc on the battlefield. Hence, there will need to be a serious alteration in military training programmes by integrating complex topics of AI and its application in the battlefield. This will also significantly reduce the number of humans on the ground when it comes to direct confrontation. Due to highly reliable and accurate intelligence of the battlefield, commanders may feel hesitant to send their soldiers when danger is imminent.
This brings the next point: currently, drones are primarily being incorporated into AI and unmanned weapons, which are in the aerial domain. It can be expected that this will have spillover effects to both land and sea with the extensive use of robots like China’s army of robot dogs. Military scientists will look to take it up a notch and coordinate the movements between these across multiple domains. For example, the drones will relay information from the air to mission control, which will then be used to guide the movements of unmanned weapons across the theatre of war.
The main question that arises is what happens to the laws of war and human rights? We have had to go through heinous atrocities over the past during wartime to make it somehow civilised. But with AI’s evolving role, who is to be held accountable if a non-sentient object starts pulling the trigger? There were many instances where humanity prevailed over the brutality of war in the past. We might never get to witness episodes of wholesomeness like the ‘Christmas truce of 1914’ ever again. For now, the countries engaged in AI warfare have stuck to ‘Human-on-the-loop’; however, as AI starts to deeply get incorporated into warfare, autonomous protocols might soon replace them.
However, with optimisation in non-combat roles, humanitarian work could be done more efficiently. Demining could be done with almost no human casualties, while early warning systems and analysis of the best evacuation routes would save innocent lives. Requirements for humanitarian aid such as food, water and medicines could be calculated faster in critical situations, along with their transport using technologies such as drones and robots.
Additionally, the interest of AI companies in getting involved in the Russia-Ukraine war has shown their growing influence. The growing dependence of states on these companies has drawn criticism, as many point out that these companies can have access to sensitive and confidential government data, which might be used in their vested interests.
Conclusion
The war in Ukraine has been a canvas for AI companies to help states revolutionise warfare and show the world that AI is inevitable. It will be used for all Combat, Combat Support and Non-combat roles. Ukraine's collaboration with private AI companies has turned warfare from the use of guns and humans towards data and algorithms. This has made war terrifying than before. Finally, the long-standing norms of international humanitarian law will need to be updated, keeping the implications of AI in mind.
