CSIS fellow analyzing Russia's AI-enabled warfare systems
Kateryna Bondar
Profile
Kateryna Bondar is a senior fellow at the CSIS Wadhwani AI Center, and she occupies a niche that almost nobody else does: writing about military AI from the position of someone who watched it get built under fire. Before Washington, she was an adviser inside Ukraine’s government, working on defense, financial-sector, and innovation reform — earlier still, project work with Ukraine’s National Reforms Council and the Ministry of Finance reform team, defense technical-assistance projects at PwC, and a stint in financial control at Microsoft. She holds bachelor’s and master’s degrees in international relations from Taras Shevchenko National University of Kyiv, and spent 2019–2020 as a fellow in the Ukrainian Emerging Leaders Program at Stanford’s Center on Democracy, Development and the Rule of Law. After Russia’s full-scale invasion she worked on intelligence-sharing coordination with Western partners and on the mechanics of G7 financial support.
What makes her worth reading is the reporting. Most “AI and the future of war” writing is a literature review with a threat model bolted on. Bondar’s reports are built on dozens of interviews with Ukrainian brigade commanders, procurement officials, and startup engineers, and they contain the kind of detail that only comes from asking operators what actually works. Her November 2024 report Understanding the Military AI Ecosystem of Ukraine mapped the whole stack — 243 AI companies, a 300,000-strong IT workforce, and the Brave1 cluster that routes frontline requirements to civilian developers. Her March 2025 follow-up on AI-enabled autonomous warfare documented something developers will recognize immediately: Ukraine took open, publicly available computer-vision models, fine-tuned them on classified frontline footage nobody else has, and shipped them onto cheap airframes — raising terminal hit rates several-fold and cutting drones-per-target from eight or nine to one or two. That is transfer learning as an artillery multiplier, and it is the single clearest real-world demonstration that data moat plus commodity model beats expensive bespoke system.
Her more recent work turns the same lens on Russia, and it is less comfortable reading. How Russia Is Reshaping Command and Control for AI-Enabled Warfare (February 2026) and How Russia Is Building a Sovereign Drone Ecosystem for AI-Driven Autonomy (April 2026) argue that Moscow has stopped trying to build a grand centralized military-AI program and instead adopted the thing that actually works: apply available tools, test in combat, scale what survives, and wire state policy, civilian industry, and frontline feedback into one loop aimed at the 2030s. The corresponding New York Times op-ed put the blunt version in front of a general audience. Her June 2026 brief with Matt Mande, Defining Autonomy, is the one to hand a software engineer: it argues that treating the drone as the “autonomous weapon” is a category error, because the lethal decisions increasingly live in the orchestration layer above the effectors — and that both Ukraine (Delta) and Russia (Svod) have fielded such layers while the U.S. has not.
For a developer, the value here is calibration. The defense-AI conversation in the U.S. is dominated by vendors — Anduril under Palmer Luckey, Shield AI under Brandon Tseng, Palantir under Alex Karp — all of whom have a product to sell and a story that fits it. Bondar has no platform to pitch. Her recurring finding is unglamorous and consistent: the decisive variables are procurement speed, data pipelines, orchestration software, and the willingness to field something mediocre this month rather than something excellent in three years. Her July 2025 report on drone dominance is essentially an argument that the Pentagon’s binding constraint is acquisition culture, not technology. Read her alongside the American vendor pitch decks and you get a much more honest picture of what “AI at war” currently means.
Key Articles & Papers
Understanding the Military AI Ecosystem of Ukraine Ukraine's Future Vision and Current Capabilities for Waging AI-Enabled Autonomous Warfare Defining Autonomy: Why Software, Not Drones, Will Decide the Next War How Russia Is Building a Sovereign Drone Ecosystem for AI-Driven Autonomy How Russia Is Reshaping Command and Control for AI-Enabled Warfare Unleashing U.S. Military Drone Dominance: What the United States Can Learn from Ukraine How Ukraine Rebuilt Its Military Acquisition System Around Commercial Technology Russia Is Building Tomorrow's War Machine What Are the Unintended Consequences of the U.S.-Iran Conflict for Defense and Security?Videos
Spotify Podcasts
YouTube
