Ukraine’s friends will be allowed to train artificial intelligence systems on real battlefield datasets gathered during the war with Russia, the Ministry of Defense in Kyiv said in early March.
Companies will receive controlled access to data specifically tailored for their needs, while the full battlefield database will remain under the control of the Ukrainian government.
Ukraine, which has millions of annotated images collected during tens of thousands of combat flights, is the first country to make such a move, the ministry said, describing it as a potentially beneficial partnership.
“Partners gain the opportunity to train their AI models on real data from modern warfare while Ukraine gains faster progress in the development of autonomous systems,” said the tech-savvy Defense Minister Mykhailo Fedorov. “We are ready to work with partners on joint analytics, model training, and the development of new technological solutions.” Fedorov was formerly the tech minister.
Kyiv is offering a strategic proposition to allies: bring your algorithms, and Ukraine will provide access to information and experience that no laboratory, simulation, or peacetime exercise can replicate. The country’s allies and suppliers have been discovering as much in the Middle East, where more than 200 service personnel have been deployed to help defend allied states against Iranian air attacks. Reports suggest they’ve been appalled by what they have witnessed.
AI is taking on a growing role. At a time when it is increasingly clear that the future of warfare belongs to autonomous systems, high-quality data is a key ingredient in training effective neural networks.
“Ukraine has accumulated what is essentially a universal military dataset,” said Ruslan Prylypko, head of command and control information systems at Aerorozvidka, the famed tech volunteer group which has helped develop Ukraine’s drone capabilities.
“The dataset contains annotated images and video of a wide range of battlefield objects, including vehicles, drones, infantry, and other targets collected during the actual deployment of unmanned systems,” he said. “Because everything is carefully tagged and classified, developers can train models for very specific tasks.”
If a company is building a system to intercept Shahed drones, for example, it doesn’t need images of tanks or trenches, but thousands of examples of Shaheds from different angles so the model can reliably recognize them, Prylypko said.
The data also reflects constant changes on the battlefield, ensuring AI is trained with the latest information. Russian equipment is often modified with cages, improvised armor, or camouflage structures, for example, so a tank may not be recognizable as a tank to an AI system trained on older information.
Updates to Ukraine’s Delta digital information network show the benefits of the program. The system, which has been deployed across all levels to support real-time battlefield awareness and operational planning, now has an AI platform that enables automatic detection of enemy equipment based on battlefield data.
Get the Latest
Sign up to recieve the Age of Autonomy newsletter and CEPA’s latest work on Defense Tech.
The new capability improves the speed with which units can identify and engage the enemy and supports the targeting of more than 2,000 Russian assets every day, the ministry said.
Access to battlefield datasets for international partners is the latest stage in a broader strategy to turn frontline experience into defense innovation.
Kyiv has worked to smooth the path for developers, including a March 2025 law enabling the transfer of defense designs to private companies for production, and November 2025 legislation on the ownership of intellectual property created during military service.
Ukraine has also helped defense industries bridge the gap between their prototypes and relevance at the front, where the innovation cycle runs in months, or sometimes weeks.
By opening access to battlefield data, Kyiv is alerting its partners to the fact that the future of war is already here, and those who want credible autonomous capabilities must train them in real war conditions.
Last year, the Ukrainian defense-tech cluster Brave1 launched the “Test in Ukraine” platform, allowing foreign developers to trial drones, robotics, AI, and electronic warfare technologies in conditions close to real combat, with direct feedback from frontline users.
In January, it followed this with the Brave1 Dataroom, a secure environment for testing and training AI models using real battlefield data. The platform contains curated visual and thermal datasets and is designed to accelerate the development of autonomous systems capable of detecting and intercepting aerial threats.
AI-enabled autonomy in warfare raises profound concerns, especially around accountability, target verification, and ethics, and not every issue has been settled. This is particularly true around the preservation of human-in-the-loop vs completely autonomous weapons.
But rather than being an argument against the initiative, it is an argument for building robust rules while the technology is being fielded.
Ukraine is not just a recipient of security assistance, but a source of military knowledge that allies need, whatever President Trump may say. The escalation of the US-Israeli war on Iran has shown there’s no substitute for battlefield experience, and Ukraine simply has more of that than anyone else.
Elena Davlikanova is a Democracy Fellow with the Center for European Policy Analysis (CEPA). Her work is focused on Ukraine and Russia’s domestic issues and their effects on global peace. She is an experienced researcher who, in 2022, conducted the studies ‘The Work of the Ukrainian Parliament in Wartime’ and ‘The War of Narratives: The Image of Ukraine in Media.’
Europe’s Edge is CEPA’s online journal covering critical topics on the foreign policy docket across Europe and North America. All opinions expressed on Europe’s Edge are those of the author alone and may not represent those of the institutions they represent or the Center for European Policy Analysis. CEPA maintains a strict intellectual independence policy across all its projects and publications.
Europe’s Edge
CEPA’s online journal covering critical topics on the foreign policy docket across Europe and North America.