The Current State of Russian Military AI

Ioannis Sidiropoulos

Russia is fielding narrow AI in drones, electronic warfare, and decision support while remaining constrained in frontier computing and fully autonomous warfare.

Russia has treated artificial intelligence (AI) as a strategic technology for several years, but the war in Ukraine has exposed a significant gap between Moscow's ambitions and its demonstrated capabilities.[2] Russia is neither an AI laggard nor a peer of the United States and China. Its comparative strength lies in rapidly integrating relatively narrow AI functions into drones, electronic warfare (EW), surveillance, and battlefield decision support rather than in frontier AI or fully autonomous warfare.[2][3]

The most important development since 2022 is therefore not a Russian “AI revolution,” but the progressive militarization of applied autonomy. The battlefield has become a testing ground where Russia is trying to make unmanned systems less dependent on continuous communications and human operators.[3]

Strategy and institutional architecture

Russia's AI policy is explicitly connected to national security and technological sovereignty. Presidential Decree No. 124 of February 2024 updated the National AI Development Strategy through 2030 and identified computing infrastructure, domestic electronics, research capacity, data and international cooperation as strategic priorities. The revised strategy sets a target of increasing the combined capacity of Russian AI-oriented supercomputers to one exaflop by 2030, compared with 0.073 exaflops in 2022.[1]

Within the Ministry of Defense (MoD), AI development has become increasingly institutionalized. In 2022 the ministry announced a dedicated AI department intended to coordinate research, development, testing and deployment across military organizations and defense companies. This sits alongside the longer-standing Russian emphasis on military robotics and unmanned systems.[2]

The institutional model is highly centralized and state-led. Defense organizations such as Rostec and the Foundation for Advanced Studies interact with universities and major civilian technology companies. This is important because Russia's military AI ecosystem is not isolated from its civilian sector: technologies developed by companies such as Yandex and Sber can potentially feed into defense applications, while military requirements increasingly influence civilian tech priorities.[2]

What Russia is actually fielding

The clearest evidence of Russian military AI is found in unmanned aerial systems. Russia has deployed enormous numbers of drones in Ukraine, ranging from relatively simple reconnaissance and attack systems to increasingly sophisticated platforms incorporating computer vision, navigation and autonomous functions.[3][5]

The Lancet loitering munition is particularly important. Its manufacturer, ZALA/Kalashnikov, has described the system as capable of autonomously locating and striking targets. The KUB-BLA has similarly been advertised with “artificial intelligence visual identification” capabilities. However, claims about the degree of autonomy should be treated cautiously. Open-source evidence does not establish that these systems routinely conduct independent target selection comparable to the most ambitious descriptions of autonomous weapons.[2][8]

More convincing evidence concerns AI-assisted navigation and terminal guidance. In October 2024, Defense Minister Andrei Belousov publicly stated that Russian forces had deployed AI-equipped drone detachments in Ukraine and nearby Russian regions. Russian military personnel demonstrated drones using automatic image recognition that could continue toward a recognized target after losing the communications link.[5]

This development is strategically significant. Electronic warfare is pervasive in Ukraine, and communications links are vulnerable to jamming. AI that allows a drone to navigate, recognize a target or continue a mission without continuous external control therefore has an immediate military utility even without sophisticated general-purpose intelligence.[3][5]

By 2026, this trend had expanded beyond attack drones. In June 2026, Belousov said Russia was developing an AI-assisted air-defense decision-making system intended to improve detection and management of Ukrainian drone attacks, with deployment planned for later that year. The project reportedly combines a unified information system with software intended to provide commanders with a more integrated operational picture.[6]

Electronic Warfare and decision support

EW may ultimately be one of Russia's more consequential AI applications. Moscow has invested heavily in EW for decades, and AI offers opportunities to automate the identification, prioritization and response to signals.[2][7]

The RB-109A Bylina, for example, has been described as an automated EW decision support system capable of identifying potential targets and selecting appropriate suppression measures. Chatham House notes, however, that many details concerning the system's actual deployment remain based on Russian reporting rather than independently verifiable battlefield evidence.[7]

This illustrates a broader problem: Russian military AI is difficult to assess from open sources. Russia routinely publicizes ambitious technological programs, while battlefield evidence frequently reveals more modest capabilities. CNA concluded that, despite extensive Russian discussion of AI before and during the war, publicly observable evidence of advanced military AI remained limited.[4]

The Ukraine war is changing the trajectory

The most important lesson from Ukraine is that Russia's military AI strategy is becoming pragmatic and incremental.

Earlier Russian thinking frequently envisioned highly automated command systems and robotic formations. The reality of the war has instead encouraged investment in technologies that solve immediate battlefield problems: reconnaissance, target recognition, navigation in contested electromagnetic environments, drone interception and rapid processing of battle information.[2][3]

The April 2026 CSIS assessment is particularly important because it argues that Russia is now moving beyond purely remote-controlled unmanned systems toward what it calls “functional independence at the tactical edge.” Its assessment is not that Russia has achieved comprehensive autonomous warfare, but that narrow AI functions are increasingly being embedded directly into battlefield platforms.[3]

This distinction matters. Russia does not need to reproduce American or Chinese frontier AI to obtain military benefits. A relatively inexpensive drone able to recognize a target or maintain navigation despite communications disruption can generate substantial tactical value.[3]

“Russia does not need to reproduce American or Chinese frontier AI to obtain military benefits.”

Constraints remain substantial

Russia nevertheless faces major structural obstacles. The first is computing and semiconductor access. Moscow's own AI strategy identifies computing capacity as a central constraint, while Western sanctions restrict access to advanced processors and other high-end technologies. Russia is attempting import substitution, domestic chip development and cooperation with non-Western partners, but these efforts do not eliminate the gap with the United States or China.[1][2]

The second constraint is human capital and innovation. Russia possesses strong mathematical and engineering traditions, but the war, emigration and restricted access to international research networks complicate the development of globally competitive AI ecosystems. Centralization can mobilize resources quickly, but it can also discourage experimentation and innovation.[2]

The third is data and integration. Military AI requires reliable, timely and interoperable data. Russia has extensive government and military datasets, but converting them into high-quality training data and integrating systems across services remains difficult.[2]

Finally, Russia faces the perennial problem of testing advanced systems under realistic conditions. Ukraine provides an unparalleled combat laboratory, but it is also an exceptionally hostile environment. Drones suffer heavy attrition, communications are disrupted, and algorithms must function amid deception, camouflage and rapidly changing tactics.[3][4]

Strategic assessment

The best assessment is therefore neither that Russian military AI is largely propaganda nor that Russia has already achieved autonomous warfare.

Russia has demonstrated a meaningful ability to industrialize narrow AI applications, particularly in unmanned systems. The strongest evidence concerns autonomous or semi-autonomous navigation, computer vision and adaptation to electronic warfare. Its 2026 efforts to introduce AI into air-defense decision-making suggest that the application is expanding from individual platforms toward broader command-and-control functions.[3][5][6]

At the same time, evidence for genuinely autonomous lethal decision-making remains much weaker than Russian rhetoric sometimes suggests. CNAS similarly concludes that Russia's AI ambitions are significant but constrained by sanctions, technological dependencies and the difficulty of translating military concepts into operational capabilities.[2]

For NATO and European defense planners, the implication is practical: the relevant Russian threat is not a hypothetical Russian superintelligence, but the mass deployment of increasingly capable, inexpensive, and partially autonomous systems. Russia can combine large-scale drone production with AI-enabled navigation, targeting, and EW. This may create saturation effects even without world-leading foundation models.[2][3][5]

“The relevant Russian threat is not a hypothetical Russian superintelligence, but mass deployment of increasingly capable, inexpensive, partially autonomous systems.”

The emerging Russian model is consequently best understood as “good-enough AI at scale.” Its effectiveness will depend less on matching Silicon Valley in frontier AI than on repeatedly translating battlefield lessons into cheap, resilient systems and producing them in quantities sufficient to overwhelm adversaries' technological and organizational advantages.[3] For European assessments and procurement, deployed resilience, iteration speed, and scale may matter more than headline model sophistication.

11. REFERENCES. USE THE CHICAGO STYLE FOR ENDNOTES AND ALSO ANY MAPS OR CHARTS. Use Calibri 12 single-spaced inside paragraphs, with 1.5 spaces between paragraphs.

[1] President of the Russian Federation, “Decree No. 124: On Amendments to Decree No. 490 and the National AI Development Strategy to 2030,” February 15, 2024, https://regulations.ai/regulations/russia-2024-2-updated-ai-strategy.

[2] Samuel Bendett, “The Role of AI in Russia’s Confrontation with the West,” Center for a New American Security, May 3, 2024, https://www.cnas.org/publications/reports/the-role-of-ai-in-russias-confrontation-with-the-west.

[3] Kateryna Bondar, “How Russia Is Building a Sovereign Drone Ecosystem for AI-Driven Autonomy,” Center for Strategic and International Studies, April 13, 2026, https://www.csis.org/analysis/how-russia-building-sovereign-drone-ecosystem-ai-driven-autonomy.

[4] CNA, “Artificial Intelligence and Autonomy in Russia: A Year’s Reflection,” September 2022, https://www.cna.org/analyses/2022/09/ai-autonomy-in-russia-a-years-reflection.

[5] Reuters, “Russia Says It Is Ramping Up AI-Powered Drone Deployments in Ukraine,” October 11, 2024, https://www.reuters.com/business/aerospace-defense/russia-says-it-is-ramping-up-ai-powered-drone-deployments-ukraine-2024-10-11/.

[6] Anadolu Agency, “Russian Defense Minister Says Moscow Developing AI System to Counter Ukrainian Drone Attacks,” June 29, 2026, https://www.aa.com.tr/en/eurasia/russian-defense-minister-says-moscow-developing-ai-system-to-counter-ukrainian-drone-attacks/3981009.

[7] Chatham House, “Military Applications of Artificial Intelligence: The Russian Approach,” in Advanced Military Technology in Russia, September 2021, https://www.chathamhouse.org/2021/09/advanced-military-technology-russia/06-military-applications-artificial-intelligence.

[8] Automated Decision Research, “Weapons Systems with Autonomous Functions Used in Ukraine,” accessed September 5, 2026, https://automatedresearch.org/news/weapons-systems-with-autonomous-functions-used-in-ukraine/.