Russia's Armored Reconstitution and AI Command Architecture: Preparing for Extended Confrontation

The Russian defense establishment is pursuing a comprehensive modernization program that extends far beyond the immediate demands of the Ukraine war, revealing strategic preparations for sustained military confrontation with NATO. Recent intelligence assessments and leaked internal documents from Russia's primary tank manufacturer demonstrate that Moscow is simultaneously rebuilding its conventional armored forces while fundamentally reshaping its command and control architecture around artificial intelligence and unmanned systems.

Internal documents published in October 2025 by Frontelligence Insight exposed Uralvagonzavod's ambitious production targets: an 80 percent increase in T-90 tank output by 2028 compared to 2024 levels, alongside the introduction of the new T-90M2 variant (also designated Project 188MS or Ryvok-1). The manufacturer expects to produce 10 T-90M2 tanks in 2026, scaling to a peak of 428 combined T-90M and T-90M2 units in 2028. Between 2027 and 2029, plans call for manufacturing 1,118 new and modernized T-90M and T-90M2 tanks. When combined with T-72B3M modernization efforts, Frontelligence assesses that Russia aims to field more than 2,000 advanced main battle tanks between 2026 and 2036, sufficient to fully restore pre-war armored strength.

These production figures represent a dramatic escalation over current capacity. Ukrainian intelligence estimates Russia's ideal annual T-90 production at 60 to 70 tanks, with monthly output between three and six units. A Russian military blogger claimed in July 2025 that the factory produced between 540 and 630 T-90M tanks since February 2022, averaging 13 to 15 tanks monthly. Meeting the new targets will require substantial industrial expansion, particularly through automation and high-precision machine tools. Despite Western sanctions, Russia continues acquiring such equipment through evasion networks. Ukraine's Main Military Intelligence Directorate reported in March 2025 that Uralvagonzavod produces tank engines using European-manufactured CNC machines obtained illegally.

Open-source satellite analysis reveals Russia is also accelerating T-72 refurbishment efforts. Between June and October 2025, Russia's overall tank reserve declined from 3,106 to 2,478 units, with T-72A stocks dropping from 900 to 461. Current estimates indicate remaining reserves include approximately 141 T-54/55 tanks, 885 T-62s, 611 T-64s, 492 T-72As, 287 T-72Bs, 44 T-80Bs, 82 T-80Us, and zero T-90s in storage. The withdrawal of T-72As from previously untouched depots and the cannibalization of T-64s for parts suggest Russia is systematically working through its Soviet-era stockpiles while simultaneously building modern replacement capacity.

Crucially, the Institute for the Study of War assesses that Russia is conserving its refurbished tanks rather than deploying them in the high-intensity armored assaults that characterized 2023 and 2024 operations. Russian forces have shifted toward lighter armored vehicles and infantry-supported attacks, likely driven by devastating Ukrainian drone effectiveness against massed armor. However, Russia still conducts occasional company- and battalion-sized mechanized assaults when weather conditions—particularly rain and high winds—degrade Ukrainian drone operations. These selective uses of armor suggest Russia is building a strategic reserve for deployment in future operations, whether in Ukraine or against NATO members.

While conventional force reconstitution proceeds, Russia is simultaneously pursuing a fundamental transformation of its command and control architecture centered on unmanned systems and artificial intelligence. This parallel modernization effort may prove more consequential than tank production for Russia's future military effectiveness. By early 2025, unmanned aerial systems reportedly conducted up to 80 percent of Russian fire missions, with forces striking approximately 300 targets daily. The central operational challenge has shifted from deploying individual drone platforms to managing heterogeneous unmanned fleets within an integrated kill chain—a problem that Russian military leadership now treats as the highest priority.

Russian military scholars acknowledged in early 2025 that despite urgent operational need, the Ministry of Defence had not fielded a fully functional large-scale UAS management system, placing Russia approximately 1.5 to 2 years behind Ukraine's Delta platform and integrated control modules. This assessment prompted a strategic reorientation away from comprehensive Automated Command and Control System (ACCS) concepts—Russia's equivalent to Joint All-Domain Command and Control—toward narrower, functionally specific software solutions addressing immediate battlefield requirements.

The Glaz/Groza software complex exemplifies this pragmatic approach. Originally designed for artillery fire adjustment, the system evolved into a comprehensive reconnaissance-strike ecosystem linking UAS operators, fire direction centers, and artillery units into a single digital workflow. The Glaz suite provides drone operators with real-time geolocation, target marking, and coordinate extraction from live footage, installed on DJI and Autel controllers as well as Android tablets. Groza functions as the fire control and mission management hub, running on Windows laptops or Android tablets with full digital mapping, automated ballistic calculations, and rapid transmission of coordinates to firing units. The system recently integrated drone mission planning capabilities to address frequency coordination and friendly jamming problems.

Available evidence suggests Glaz/Groza entered pilot testing with combat units in early 2024, achieving relatively widespread deployment by August 2025 across drone units, artillery batteries, and reconnaissance elements. The integrated workflow reduces time from target detection to artillery impact from hours to minutes. A UAS operator places a crosshair on a target, Glaz extracts coordinates from telemetry, Groza performs ballistic calculations and transmits fire missions, and subsequent corrections flow through the same digital loop without voice reports or manual orientation procedures.

Russia's broader command modernization effort includes systematic data infrastructure development supporting AI model training. Defence Minister Andrei Belousov initiated this effort in mid-2025 by directing establishment of a unified database recording enemy losses inflicted by drones. A video from August 2025 showed Ministry of Defence staff demonstrating a system that automatically aggregates and analyzes operator-reported drone activity, linking each operator's performance trajectory to unique personal identification numbers. This eliminates opportunities to manipulate results while creating comprehensive training datasets. In September 2025, the ministry convened a Technical Council on Development of the Drone Management System, bringing together senior military leaders, research institutes, and operational units to advance unified UAS command architecture.

Russian military doctrine assigns AI two core functions: enhancing sensor data processing and providing predictive decision support through scenario generation and recommendations. Critically, AI is envisioned as augmenting rather than replacing human commanders, who retain formal authority and responsibility. Russian military assessment places computer vision, sensor fusion, and signal analysis capabilities at technology readiness levels 6 through 9—relatively mature and field-testable. By contrast, natural-language processing remains at experimental stages, TRL 1 through 3, reflecting deliberate prioritization of applications with immediate battlefield utility where abundant combat data enables rapid validation.

The Platform-GNS software environment, developed by the Center for Artificial Intelligence Technologies at Zhukovsky Research Center, provides the infrastructure for this AI development. Platform-GNS offers a complete technological stack for building applications based on deep convolutional neural networks, supporting the entire lifecycle from dataset preparation through model training, testing, and deployment. The specialized Platform-GNS Avtomat variant focuses specifically on high-precision ground target recognition from airborne sensors. Both systems are distributed free to Russian defense enterprises, Ministry of Defence organizations, and educational institutions, facilitating rapid adoption. Planned 2025-2026 upgrades will add large language model support for smart assistants and more autonomous systems.

The January 2026 announcement of Russia's next-generation weapons rollout provides additional context for these modernization efforts. The Su-75 Checkmate fifth-generation fighter is beginning test flights, intended to replace aging MiG-29s with a supersonic, stealthy platform carrying significant export potential. The Khabarovsk nuclear-powered submarine prepares for sea trials, designed to carry six Poseidon nuclear-powered underwater drones for strategic deterrence. The Sarmat heavy ICBM enters final flight tests to replace the Voevoda missile, capable of carrying multiple ballistic or hypersonic warheads. Zircon hypersonic missile production is ramping up following operational testing in Ukraine, with deployment planned for Yasen-M submarines and the Admiral Nakhimov cruiser. The S-500 Prometheus anti-ballistic missile system continues rolling out to defend critical sites against near-space threats.

These parallel streams of modernization—conventional force reconstitution, AI-enabled command architecture, and next-generation strategic systems—reveal an integrated approach preparing Russia for extended confrontation. The timeline for threat materialization may be shorter than Western assessments suggest. Russian forces do not require complete restoration of pre-2022 strength before initiating new operations if NATO deterrence appears insufficient. Moreover, Russia is already engaged in what the Institute for the Study of War characterizes as Phase Zero operations: sabotage, reconnaissance, and irregular activities targeting European critical infrastructure and military assets.

Recent drone incursions near NATO facilities support this assessment. Czech military observers reported increased drone activity near military facilities in October 2025. On October 8, an unidentified drone penetrated the security perimeter of the NATO air base at Geilenkirchen, flying at low altitude across the runway hosting airborne early warning and control aircraft essential to eastern flank surveillance. While German authorities did not formally attribute the incident, the pattern aligns with broader Russian condition-setting behavior preparing the operational environment for potential future escalation.

Russia's draft budgets for 2026 and 2027 project continued defense and national security spending around 8 percent of GDP. Some observers question whether Russia can sustain such expenditure levels, particularly given constraints on equipment production and sanctions limiting critical component imports. However, Russian military leadership has demonstrated consistent ability to absorb failures, adapt under pressure, and scale solutions that prove operationally effective. The shift from comprehensive ACCS concepts to task-specific battlefield software, the rapid institutionalization of volunteer-developed tools like Glaz/Groza, and the systematic creation of AI training datasets all reflect this pragmatic adaptation.

The fundamental challenge facing NATO is that Russia's preparation for extended confrontation proceeds across multiple dimensions simultaneously, creating overlapping timelines for threat development. Conventional armored reconstitution requires years and substantial industrial investment, but AI-enabled command improvements, drone swarm tactics, and asymmetric capabilities can mature more rapidly. Russia's parallel development of long-range strike capabilities, underwater nuclear drones, and hypersonic missiles provides immediate strategic deterrence independent of tank production schedules. The combination creates a force that may achieve operational effectiveness through technological sophistication and tactical innovation rather than numerical superiority in legacy platforms.

European responses demonstrate growing recognition of these threats. Ukraine continues expanding defense industrial partnerships, including a memorandum with the Netherlands for joint deep-strike drone production with €110 million in funding, €200 million for air defense systems, and agreements with the United Kingdom for artillery production and the LYRA battlefield technology program. These initiatives aim to enhance long-term defense capacity while bolstering NATO-Ukraine industrial cooperation.

The broader strategic implication is that Russia's military posture reflects deliberate preparation for conflicts extending beyond current operations in Ukraine. The expanding tank production, sophisticated command architecture development, and ongoing reconnaissance activities in Europe collectively indicate Russia is building a flexible, resilient force designed to project power and deter adversaries across multiple domains. NATO planning must account not only for conventional force regeneration timelines but also for Russia's capacity to leverage combat experience, revised doctrine, and rapidly maturing autonomous systems in future confrontations. The operational and industrial measures now underway suggest Russian military threats may intensify following any cessation of hostilities in Ukraine, making continuous assessment and adaptive deterrence essential priorities for the alliance.