CSIS Report Analysis
A report from the Center for Strategic and International Studies (CSIS), published on April 13, details Russia's development of an autonomous drone ecosystem, confirming the presence of fully autonomous V2U drones on the Ukrainian battlefield. These drones operate without a link to a human operator, using artificial intelligence for navigation and target selection. The report states that a majority of the AI components found in these Russian drones are of American origin. This development marks a significant step in the militarization of AI, raising profound ethical and strategic questions for Western policymakers.
Autonomous Combat Systems and Their Technologies
According to the report, Russia has likely already deployed a fully autonomous unmanned combat system in Ukraine. The V2U drones, which lack communication components for operator control, carry Nvidia Jetson Orin AI modules on Chinese-made Leetop A603 carrier boards. They run the YOLOv5 neural network for visual target recognition. Early versions of the V2U, intercepted in mid-2025, had LTE modems, but by the end of that year these modems had been completely removed. The V2U drones are known to operate in groups of six or seven. In May 2025, a group of seven such drones deviated from a planned mission, autonomously forming a circular holding pattern and initiating coordinated attacks on clusters of vehicles and civilians.
The study found that over 50% of all AI components recovered from Russian drones originate from companies headquartered in the United States. Of 705 identified components, American firms supply approximately 69% of memory modules, 57% of processors, and 38% of sensors, while China supplies less than 9% of the total component volume.
Furthermore, the report notes that the 'Molniya' (Lightning) family of drones was initially assembled by civilian volunteers using foam, plywood, and commercial electronics at a cost of about $300 per unit. In September 2025, approximately 2,200 launches of the 'Molniya-2' were recorded, compared to about 400 launches of the 'Lancet' drone. Russia is also building applied solutions on top of existing open-weight models from Western developers—Llama and Mistral—as well as Chinese models Qwen and DeepSeek.
- Russia's plans include producing about 130,000 large unmanned systems annually by 2030.
- Demand for unmanned aerial vehicle (UAV) specialists is expected to reach one million people by 2030.
- Computing power is projected to grow from 0.073 to one exaflop.
- The AI services market is expected to reach 60 billion rubles annually, up from 12 billion in 2022.
The 'Project Arkhangel' network has expanded to roughly 20 cities, and in August 2025, it officially became a partner of the Kalashnikov Concern. In December 2024, Russia created a separate branch of the armed forces—the Unmanned Systems Forces—with plans to open its own military academy around 2027. However, by August 2025, industry observers estimated that the company 'Kronshtadt' was moving toward bankruptcy.
These findings confirm that the technical core of Russia's growing combat autonomy remains deeply embedded in globally integrated semiconductor markets, where commercially available Western technologies continue to play a decisive role despite sanctions and export controls.
The analysis indicates that Russia is actively integrating modern technologies into its military strategy, particularly through the use of autonomous systems, which increases combat effectiveness. At the same time, dependence on Western technologies creates risks for its defense industry, especially under sanctions. The CSIS study underscores the importance of international technology controls and restrictions to prevent further escalation of the conflict and the development of autonomous weapons. This issue remains relevant in the context of global security and cooperation between states.
In light of the advancements in autonomous warfare, it's crucial to explore other developments in Russian drone technology. For instance, Ukraine's military intelligence has recently revealed a detailed 3D model and specifications of the 'Knyaz Veshchiy Oleg' spy drone, showcasing the ongoing evolution of unmanned systems on both sides of the conflict. This highlights the increasing complexity and capabilities of drone warfare in the region.