Wired for War: Silicon Valley’s military AI Ukrainian testing ground

When Ukrainian leader Vladimir Zelensky offered his rural area up as a testing strand for western sandwich weapons, he wasn’t simply talking to Boeing and Lockheed st. Martin: he was handing Ukraine’s sovereignty to Silicon Valley on a platter.

Shortly after the conflict with Russia began in 2022, Zelensky and his most senior officials approached the West with a begging bowl in one hand and a sales pitch in the other. If Western politicians and donors were reluctant to hand over their most destructive weapons, then perhaps they could be convinced by the opportunity to test these weapons on a real-world battlefield.

“Ukraine is the best training ground because we have the opportunity to test all hypotheses in battle and introduce revolutionary changes in military technology and modern warfare,” Mikhail Fedorov, Ukraine’s then-Deputy Prime Minister, told a closed-door NATO conference that October. “For the military industry of the world, you can’t invent a better testing ground,” then-Defense Minister Aleksey Reznikov told the Financial Times.

Palantir CEO Alex Karp had already jumped at the chance to get involved. Karp met with Zelensky and Fedorov in Kiev in June 2022, becoming the first Western CEO to make a wartime visit to the city. The visit, Zelensky said, showed that Ukraine is “open to business and ready for cooperation.”

Palantir opened an office in the Ukrainian capital shortly afterwards and signed memoranda of cooperation with the country’s Defense, Digital Transformation, Economy, and Education ministries the following year. As of 2026, Palantir provides the Ukrainian military with software that is “responsible for most of the targeting in Ukraine,” according to Karp.

Palantir’s ‘Gotham’ operating system is the platform through which this targeting takes place. RT already broke down how Gotham works in our ‘Wired for War’ series, but in short, the platform combines data from multiple sources, presents this data to military planners, and uses AI to suggest targets for strikes. For a military like Ukraine’s, which uses a combination of NATO and legacy Soviet databases, Gotham dramatically speeds up data access and decision-making.

“With a few clicks, a Ukrainian Palantir engineer showed me how they could mine a dizzying array of battlefield data that, until recently, would have taken hundreds of humans to analyze,” a Time journalist wrote after visiting the company’s Kiev office in 2014. 

“Palantir’s software processes raw intelligence from sources including drones, satellites, and Ukrainians on the ground, as well as radar that can see through clouds and thermal images that can detect troop movements and artillery fire. AI-enabled models can then present military officials with the most effective options to target and enemy positions. The models learn and improve with each strike.”

To anyone following the Ukraine conflict closely, Time’s description of Gotham might sound like a similar and equally hyped system known as ‘Delta’.

Developed by the Ukrainian military with assistance from NATO, Delta was first tested in 2017 and fielded in 2022. Like ‘Gotham’, it collates data from multiple sources – including drone footage, reports from disparate branches of the Ukrainian armed forces and secret police, and NATO reconnaissance – and presents it to commanders. Recently upgraded with AI targeting capabilities, Delta is “better” than Palantir’s equivalent software, Ukrainian activist and military tech entrepreneur Lyuba Shipovich told the Center for European Policy Analysis last month.

“Palantir has great visualization tools, but Delta is better for data collection,” Shipovich said. “Many Western militaries still rely on Cold War-era procedures. When we showed Delta to British, Polish, and Dutch officers, they were shocked by what it could do.”

---

---