Modern deep battle: The rebirth of Soviet doctrine

The Tukhachevsky era

Imagine the dust of the 1930s Soviet steppes. Men like Mikhail Tukhachevsky and Vladimir Triandafillov were not looking at the trench in front of them. They were looking at the entire nervous system of a nation. They dreamed of Deep Battle (Glubokaya operatsiya), a theory that did away with the bloody, linear slog of the First World War. Instead of punching a single hole in the line, they envisioned motorized infantry, armor, and aviation hitting the tactical front, operational reserves, and strategic rear simultaneously. It was a symphony of engineered chaos designed to induce total systemic collapse - to shatter the enemy's brain, guts, and spine all at once.

The tragedy is that neither architect lived to see the full flowering of their ideas. Triandafillov was killed in a plane crash in 1931. Tukhachevsky was executed during Stalin's Great Purge in 1937, his doctrine temporarily buried alongside him. Yet the ideas survived. And nearly a century later, they are being executed with a precision that the original theorists could never have imagined.

Wired for War: The Robotics Revolution and Conflict in the 21st Century - P.W. Singer

A foundational analysis of how robotics and autonomous systems are transforming the battlefield - covers drones, AI targeting, and the ethical dilemmas of remote warfare.

Buy now

The scale has changed, and the tools are far more precise, but the fundamental impulse to strike deep remains the primary driver of modern kinetic conflict. We are witnessing a digital resurrection of Soviet doctrine - one where the 'depth' of the battlefield is no longer just kilometers of dirt, but the very infrastructure that keeps a modern society functioning.

The drone as a scalpel in the depth

In the early hours of April 23, 2026, the air across vast stretches of contested and Russian - held territory hummed with the sound of small engines. Ukrainian forces launched a coordinated wave of long-range drones in an operation bearing clear hallmarks of Deep Battle thinking. This was not a localized skirmish - it was a strike against sustainment infrastructure spread across the full operational depth.

In the Kstovo district of Nizhny Novgorod Oblast, the Gorky oil pumping station - a critical node in Russia's Transneft pipeline network - was struck and caught fire. Simultaneously, an oil depot in occupied Feodosia, Crimea, was reportedly hit multiple times. In Melitopol, a power substation was taken offline. By targeting energy and logistics nodes across the depth of the theater, the operation was designed not to capture a single trench, but to degrade the capacity to move, sustain, and fuel armored forces across a vast front.

The numbers behind this sustained campaign are significant. Ukrainian Commander-in-Chief General Oleksandr Syrskyi reported in March 2026 that Ukrainian forces verified strikes against over 150,000 targets that month alone - a 50% increase over February - including 143 critical logistics facilities and depots and 52 command posts. Ukraine's Unmanned Systems Forces (USF) are conducting over 11,000 combat missions daily, applying persistent pressure across the full operational depth of contested territory.

Silicon strategy: AI and the compression of time

The original Soviet theorists were limited by the speed of human communication. Orders took hours or days to filter back from the front, creating a lag that blunted the intended simultaneity of deep operations. That delay has now been dramatically compressed.

The Maven Smart System (MSS), developed and operated by Palantir Technologies, represents the most publicly documented example of AI-accelerated deep operations in recent conflict. The system is built on the foundations of the US Department of Defense's Project Maven - a programme originally launched by the DoD in April 2017. Google served as the initial technology contractor, providing machine-learning capabilities, but withdrew from the programme in 2018 following significant internal employee opposition on ethical grounds. The DoD subsequently worked with alternative vendors, and Palantir Technologies emerged as the primary contractor, developing what became the Maven Smart System.

During Operation Epic Fury - the US-Israeli military campaign against Iran that ran from February 28 to April 8, 2026 - the MSS processed a torrent of data from satellites, radar, and video feeds. Using computer vision to identify targets, verify them against threat databases, and present actionable intelligence to commanders, the system compressed the kill chain from hours to minutes. On the opening day of the operation, over 1,000 targets were struck. By the time a ceasefire took hold on April 8, the White House confirmed more than 13,000 AI-assisted strikes. This is the operational realization of Deep Battle at its most advanced: the ability to strike an adversary's entire territorial depth with a speed that outpaces the capacity to react, regroup, or even fully assess the scale of damage until it is irreversible.

The ethical frontier of AI-assisted targeting

The compression of the kill chain raises profound questions that military doctrine has yet to fully answer. When an AI system can identify, verify, and present a target in minutes rather than hours, the window for human judgment narrows accordingly. The laws of armed conflict - including the principles of distinction, proportionality, and precaution - were written for a world where the pace of targeting allowed time for deliberation. The Maven Smart System is designed to keep a human decision-maker in the loop, but critics and legal scholars argue that the operational tempo generated by such systems effectively reduces meaningful human control to a procedural formality rather than a substantive check.

The Kill Chain: Defending America in the Future of High-Tech Warfare - Christian Brose

Written by a former Senate Armed Services Committee staff director, this book examines how AI and networked systems are reshaping military decision-making and the kill chain.

Buy now

Academic and legal debate on Lethal Autonomous Weapons Systems (LAWS) has intensified alongside the operational deployment of AI-assisted targeting platforms. The International Committee of the Red Cross (ICRC) has called for new legally binding rules governing autonomous weapons, arguing that human involvement in targeting decisions must be substantive, not symbolic. The tension between operational advantage - the speed and scale offered by systems like the MSS - and the ethical obligation to ensure accountable, deliberate targeting is one of the defining strategic dilemmas of the current period. It is a question that military establishments, legal institutions, and civilian governments in multiple countries are actively wrestling with, and one that the accelerating pace of AI capability is forcing toward resolution faster than international governance frameworks can comfortably accommodate.

The friction of reality: attritional responses to deep operations

While the theory of Deep Battle is elegant, its execution collides persistently with the friction of a determined, technologically capable opponent. In the Ukrainian theater, both sides have encountered the limits of rapid deep maneuver under conditions of persistent surveillance and precision strike.

Russian forces have encountered a particular form of logistical vulnerability. Modern precision warfare means that large concentrations of fuel or ammunition - the very infrastructure required for rapid deep maneuver - are detectable and destroyable before they can be committed. The result has been a departure from the high-speed breakthroughs envisioned by the original doctrine, replaced by a slower, more attritional approach relying heavily on massed artillery, long-range glide bombs, and cruise missile strikes against energy infrastructure and industrial capacity deep inside Ukrainian-held territory.

These deep strikes reflect the same doctrinal logic - disrupting sustainment and degrading industrial capacity - but without the maneuver tempo that makes Deep Battle truly decisive. The front lines themselves have remained largely static for extended periods. Both sides are, in effect, conducting parallel deep-strike campaigns against each other's rear areas through available long-range systems, while the tactical front grinds forward in increments measured in hundreds of meters.

The lesson is one the original theorists understood implicitly: doctrine is a framework, not a guarantee. The friction of a well-supplied, technologically adaptive opponent can absorb even the most sophisticated deep operations and convert them into a prolonged war of attrition.

Electronic warfare and the drone ecosystem

No account of modern deep operations is complete without addressing the electromagnetic dimension. The drone, for all its operational versatility, is a radio-frequency-dependent system. It navigates via GPS signals, receives command inputs over radio links, and in many cases transmits targeting video through channels that are encrypted but detectable. Electronic warfare (EW) - the active jamming, spoofing, and interception of these signals - has become the primary counter-system deployed against drone operations across the conflict.

Russian EW platforms, including the Krasukha-series and Pole-21 systems, have demonstrated the ability to spoof GPS coordinates, causing drones to deviate from programmed flight paths or revert to degraded autonomous modes. Ukrainian forces have similarly developed and deployed EW capabilities to disrupt Russian reconnaissance UAVs. The result is a layered electromagnetic arms race running parallel to the kinetic conflict - a contest of signal and counter-signal that shapes whether any given mission succeeds or fails before the first strike is ever launched.

Ghost Fleet: A Novel of the Next World War - P.W. Singer & August Cole

A techno-thriller grounded in real-world research, depicting near-future conflict driven by AI, electronic warfare, and autonomous systems - cited by actual military planners.

Buy now

Long-range strike missions are partly a response to this electromagnetic environment. As closer-range drone operations encounter increasing EW saturation near active front lines, operators on both sides have adapted toward extended-range missions that push beyond the effective jamming radius of ground-based platforms, and toward AI-assisted navigation architectures that maintain course integrity even under GPS denial or degradation. It is another layer of real-time doctrinal adaptation - the deep strike evolving continuously to overcome the defenses arrayed against it.

The economic shrapnel

The effects of deep strikes do not stop at the physical target. They propagate outward through logistics networks, energy markets, and global supply chains with a reach that the Soviet theorists of the 1930s could not have anticipated. Data from the April 23, 2026 HCOB Flash PMI for France illustrates this secondary effect with unusual clarity: clients were frontloading orders, driving manufacturing output to a 50-month high as businesses rushed to secure goods ahead of expected shortages and price increases. The proximate causes included both the disruption of energy infrastructure through deep strikes and the closure of the Strait of Hormuz in the context of the Iran conflict.

The disruption of an adversary's sustainment - a core tenet of Deep Battle - now generates cascading effects that ripple through the global economy. When strikes on energy infrastructure reduce pipeline throughput, or when conflict-driven closures remove key maritime trade lanes, the cost is not borne solely by the belligerents. A manufacturer in Lyon, a logistics hub in Rotterdam, or an assembly plant in South Korea absorbs the shock in the form of rising input costs, extended lead times, and supply uncertainty. The 'depth' of the modern battlefield, in economic terms, has become genuinely global.

The future of the deep strike

We are entering an era where the distinction between 'front' and 'rear' has effectively dissolved. When a drone operator can strike a target 1,000 kilometers distant, the entirety of an adversary's territory falls within the operational zone. The persistent pressure generated by over 11,000 daily combat missions represents a level of operational tempo that the Soviet theorists of the 1930s could only sketch in outline.

As AI continues to refine target identification, prioritization, and battle damage assessment, the limiting factor in deep operations will increasingly be not finding the enemy, but managing the sheer volume of targets that a system like Maven can generate faster than human decision-makers and legal reviewers can process. This creates a new category of operational risk: the possibility that the speed of AI-assisted targeting outpaces the institutional capacity for the deliberate, accountable judgment that international law requires.

Deep Battle is no longer a historical curiosity in dusty military textbooks. It is the living architecture of 21st-century conflict - networked, accelerated by silicon, and constrained, imperfectly, by the ethical and legal frameworks that human societies are still scrambling to construct around it. The steel rain is falling. It is smarter than ever before. And the question of who controls it, and how, has never been more consequential.