A natural ally: Artificial intelligence is becoming an increasingly important tool in warfare, but it requires access to an extensive civilian digital infrastructure

The Russian-Ukrainian conflict continues to be a peculiar mixture of traditional and modern warfare techniques. Alongside drones, OSINT, cyber attacks, and precision weapons, AI (Artificial Intelligence) has emerged as a game changer, playing an increasingly critical role in the war in Ukraine. However, we are not referring here to the creation of complex centralised systems incorporating AI, but rather its distributed use. AI technologies integrate a huge flow of public and commercial information, providing answers to critical military questions. First and foremost, this includes maps with up-to-date information on the location and movement of enemy troops, evaluation of optimal weapon types, automatic analysis of intercepted enemy communications, and so on. These technologies make it possible to dispel the ‘fog of war’ and provide a significant advantage on the battlefield. NATO has long recognised the critical role of AI in future conflicts and has been working on its implementation. However, the Russian army is limited in its ability to capitalise on these advantages, primarily as a result of its lack of access to international technological and digital infrastructure.

One of the market leaders in artificial intelligence technology is the American company Palantir, which specialises in developing software for collecting and analysing big data. Palantir works closely with US government agencies, including the CIA, which is one of its investors. Several years ago, US customs successfully used the company's artificial intelligence-based algorithms to track the movements of illegal migrants and refugees (a heavily criticised move). According to an article from The Washington Post, since last year, Palantir has provided its AI solutions to the Ukrainian army for use in battlefield analysis, timely troop supply, and offensive planning.

According to the newspaper, Ukraine's success in retaking Kherson was aided in part by the use of Palantir's AI-based technological solutions. These provided accurate and reliable intelligence to the Ukrainian army, including schematics of Russian troop movements, detailed information about their locations, coordinates of weaknesses in their defences, etc. This enabled the Ukrainians to repel Russian counterattacks and breach their defences.

What makes this system truly revolutionary is that it automatically seeks out data from commercial companies to analyse the current situation on the battlefield. It uses an artificial intelligence-based tool called MetaConstellation. This tool enables Ukraine and its allies to see the commercial data that is currently available for a specific area of the battlefield. Commercial companies, it turns out, are in possession of a surprisingly wide range of helpful information: conventional optical imagery, synthetic aperture radar imagery (which is imagery that is captured through clouds), thermal imagery (which is required to detect artillery and missile fire), along with other valuable data. According to The Washington Post, the primary suppliers of this data are Maxar, Airbus, ICEYE, and Capella, as well as various scientific institutions that conduct atmospheric research.

In the case of Kherson, Palantir assessed data collected by approximately 40 commercial satellites which pass over the city every day, although in principle, even a dozen would gather sufficient data. With more data, however, it is possible to obtain images of the terrain with a focus of up to three square metres. Ukrainian officers can then use handheld tablets to request real-time and high-resolution information about the terrain and enemy positions. Therefore, the availability of high-speed, high-quality Internet is a critical prerequisite for the successful implementation of this technology. The Ukrainian military famously uses Starlink Internet, powered by 2,500 satellites in low-earth orbit.

According to a review by the information platform TechInformed (TI), automated analysis of voice messages is another example of the use of AI for military purposes. Primer, a Ukrainian software company, has modified its voice-to-text software to process intercepted Russian military messages. The updated version of this software automatically and almost instantly selects and outputs critical information relating to combat operations. The Ukrainian military no longer needs to spend hours listening to and analysing intercepted conversations. A further example is the use of advanced artificial intelligence software, Clearview AI, to process images, recognise the faces of Russian soldiers, and identify enemy soldiers who have been killed. Based on a single portrait photo, the app is able to connect to social networks and automatically retrieve profiles and a plethora of other information about a person.

According to The Washington Post, the Russian military has also attempted to use AI in its operations in Ukraine, specifically to create digital battle maps based on commercial data from Chinese satellites. However, these efforts have been hampered by a lack of sufficient and high-quality data, as well as a lack of high-quality, high-speed Internet access, which deny Russian troops real-time updates. Further, the use of AI necessitates the redistribution of tactical decision-making authority, which is incompatible with the administrative practices of the Russian military command.

Military experts believe that artificial intelligence will play a significant role in future conflicts. This was notably argued by Nikos Lutas, head of NATO's Data and AI Policy Unit, at last year's annual AI conference in London. As a result, months before Russia’s full-scale invasion of Ukraine, in October 2021, NATO defence ministers approved the Alliance's first-ever AI strategy. This strategy stresses the need to create an environment conducive to the implementation and encouragement of AI development for defence and security purposes, outlines the methods needed to accelerate AI adoption and innovation, and safeguards AI technologies against malicious use by state and non-state actors.