Palantir discusses AI with Ukraine

Military AI is easiest to misunderstand when it gets described as a magic weapon. It usually is not. It is software glued to sensors, maps, comms, and human operators — and the hard part is making that whole stack work under fire. That is why the May 12 meeting in Kyiv between Ukraine’s defence minister, Mykhailo Fedorov, and Palantir CEO Alex Karp matters. The point was not a shiny demo. The point was whether Palantir’s data and decision software can plug into Ukraine’s already messy, fast-moving war machine. ### Why Palantir? Palantir is not mainly a drone company or a weapons maker. Its value is in stitching together information from different places so commanders can actually use it. In Kyiv, Fedorov and Karp discussed AI, data analysis, and scaling tech for air defence, drone warfare, and deep-strike planning — basically the software layer that helps turn scattered battlefield inputs into decisions. (mod.gov.ua) ### What does Ukraine want from that? Ukraine laid out three priority domains — air, land, and the economy. Air came first. Fedorov said the goals are full detection of aerial targets, at least 95% interception, AI tools to predict and counter threats, and pushing the war back onto Russian territory. He also said Russia used around 27,000 Shahed-type attack drones during the winter period. That number tells you the scale problem. Ukraine is not trying to win one interception. (mod.gov.ua) It is trying to industrialize air defence decisions. ### Where does AI already show up? Not just in planning rooms. Ukraine’s defence ministry says its own AI tools are already used for combat data analytics, Russian target detection, mission planning, and improving operational effectiveness. It also showed Palantir the Army of Drones system and DELTA’s Mission Control tools, which are meant to change how units are managed. So this was less “should we use AI?” and more “which parts of the stack should Palantir help scale?” (mod.gov.ua) ### What is the clearest frontline example? The clearest example is lower-tier air defence. On May 9, Ukraine said more than 10 military units were already using a compact AI-powered turret built by a Brave1 participant. The system autonomously detects, tracks, and calculates the flight path of enemy drones, while a human confirms the shot with one button. That matters because fibre-optic drones are resistant to electronic warfare jamming, so kinetic interception becomes one of the few workable options. (mod.gov.ua) ### So is this about autonomy? Partly, but the catch is that autonomy here looks narrow, not sci-fi. These systems are doing bounded tasks — track this drone, predict this path, help this operator. The bigger military value comes from data fusion. A turret is useful. A turret connected to warning networks, mission software, and command systems is much more useful. That is exactly the kind of integration problem Palantir likes to solve. This last point is an inference from what both sides highlighted. (mod.gov.ua) ### Why bring up Shield AI? Because money is chasing the same thesis. In March, Shield AI announced a $1.5 billion Series G at a $12.7 billion post-money valuation, plus $500 million in preferred equity financing. The company framed that raise around software, simulation, and autonomy — not just hardware. Different company, same direction of travel: defence buyers want AI that can be trained in simulation, refined in use, and dropped into real operational workflows. (mod.gov.ua) ### What is Ukraine really testing? Basically, whether wartime pressure can compress the whole defence-tech cycle. Ukraine already has drones, robotic systems, and software platforms across brigades. Now it is trying to connect them into a faster loop from sensing to decision to action. If Palantir gets deeper into that loop, the story is less about one company entering Ukraine and more about Ukraine becoming the proving ground for deployable military AI. (shield.ai) ### Bottom line? This is a systems story. Ukraine is showing what defence AI looks like when it leaves the lab — human-in-the-loop turrets, target analytics, mission software, and air-defence coordination under constant drone attack. Palantir’s role, if this expands, will be to make those pieces talk to each other fast enough to matter. (mod.gov.ua) (landforces.mil.gov.ua)