DECODING: Asgard, the Ukrainian AI that cuts the planning cycle from 72 hours to 60 minutes

The 72-hour problem — and Ukraine's solution

In conventional military doctrine, a corps-level operational planning cycle — the process by which a large formation identifies objectives, assigns assets, calculates timing, coordinates logistics, and generates an executable plan — takes approximately 72 hours. This is not bureaucratic inefficiency. It reflects the genuine complexity of synchronizing hundreds of elements across a large formation, under uncertainty, against a dynamic adversary. 72 hours is fast, by traditional standards.

Ukraine's Asgard AI-assisted battle management system has reportedly reduced that cycle to 60 minutes. Not 60 hours — 60 minutes. If that figure holds under operational scrutiny, it represents not a marginal improvement in military planning speed but a fundamental change in the tempo at which large formations can operate. An army that can plan and execute at 60-minute cycles operates in a different tactical universe than one that requires 72 hours.

What Asgard promises: 10x more daily strikes

The developers behind Asgard have made a specific and audacious claim: the system enables 10 times more daily target strikes than the planning processes it replaces. This multiplier is derived from the compression of planning time — more cycles per day means more opportunities to generate and execute strike packages, more adaptation to changing battlefield conditions, and more responsiveness to emerging targets of opportunity. Speed of decision translates directly into operational intensity.

These claims are extraordinary by the standards of any military planning system. They deserve both attention and scrutiny. The following sections examine what Asgard actually does, what advantages it derives from Ukraine's unique data environment, and what limitations and risks it faces — including from Russian countermeasures.

72 hours to 60 minutes. That compression sounds like a marketing claim until you think through the implications. In 72 hours, the battlefield changes. Targets move, logistics shift, weather conditions evolve, enemy intentions become clearer or murkier. A system that plans in 60 minutes is responding to the battlefield that exists now — not the battlefield that existed three days ago. That difference is not small. It might be decisive.

A drone-native battle management system

Asgard is an AI-assisted combat management system developed by a Ukrainian defense technology team specifically for the conditions of modern warfare in Ukraine. Unlike legacy battle management systems designed for conventional combined-arms warfare, Asgard was built from the ground up for a battlefield dominated by drones, electronic warfare, and distributed sensor networks. It is, in the terminology of its developers, a "drone-native" system — one that assumes drones are primary rather than supplementary assets.

The system integrates data from multiple sources simultaneously: Bayraktar TB2 drones, commercial FPV and ISR drones, Maxar and Planet Labs commercial satellite imagery, ground-based radar, electronic intelligence (ELINT), and human intelligence (HUMINT). It aggregates, correlates, and analyzes these disparate data streams in real time, generating a fused operational picture that would take human analysts hours or days to produce manually. The AI component does not make decisions — it processes information at machine speed and presents structured recommendations to human commanders who retain decision authority.

The strike planning module

Asgard's strike planning module is its most operationally significant component. When a human operator identifies a potential target or assigns a target category, the module automatically: identifies the highest-priority targets matching the criteria, geolocates them within the operational picture, calculates optimal strike parameters (weapons selection, timing, approach vectors, deconfliction with friendly forces), and generates a complete strike plan ready for commander approval. The human-in-the-loop is preserved — no strike executes without explicit human authorization.

This workflow represents a radical compression of what has traditionally been a manual, sequential process requiring coordination across intelligence, planning, fire support, and operations staff elements. Asgard runs all of these processes in parallel, computationally, and presents the output in a format ready for commander review rather than raw staff analysis.

Human in the loop. I want to flag that phrase because it matters enormously. Asgard does not make the decision to strike. A human commander does, every time. The AI presents options, analyzes data, calculates parameters — but the finger on the trigger, metaphorically, is always human. In a war where the laws of armed conflict already face severe strain, that distinction is not a technical footnote. It is the ethical architecture of the system.

Why Ukraine's training data is unique in the world

The most significant competitive advantage Asgard possesses over any comparable system being developed in peacetime is its training data. The system was trained on four years of real combat data from one of the most intense combined-arms conflicts since World War II. This data includes strike outcomes, target movement patterns, Russian defensive responses, electronic warfare signatures, logistics flow patterns, and the thousands of other variables that define the operational environment in eastern Ukraine.

No simulation, no exercise, no wargame can replicate this dataset. Western military AI systems developed in peacetime are trained on models of war — Asgard is trained on war itself. This distinction is not philosophical — it is the difference between a medical AI trained on case studies and one trained on a decade of real patient outcomes. The latter is simply more accurate, in ways that compound as the system continues to learn from ongoing operations.

Integration with DELTA: the situational awareness backbone

Asgard does not operate in isolation. It integrates with DELTA, Ukraine's primary battlefield situational awareness platform — a system that aggregates and displays real-time information from across the front line, used by Ukrainian commanders at multiple levels. DELTA provides Asgard with a continuously updated operational picture; Asgard's planning outputs feed back into DELTA, creating a closed loop between situational awareness and operational planning.

This integration also connects Asgard to field reporting applications used by front-line units — allowing observations from the lowest tactical levels to flow into the AI's analytical pipeline. A squad leader who spots a Russian logistics column can report it through a field app; that report enters DELTA, gets incorporated into Asgard's operational picture, and potentially triggers a strike planning cycle. The network effect of these integrations is what makes the 60-minute cycle claim at least plausible.

DELTA gives Asgard the picture. Field apps give DELTA the ground truth. Asgard gives commanders the plan. And the plan goes back through the same network to execute. That is a closed-loop military decision cycle operating at machine speed. I do not know of any other military in the world that has built this at operational scale — let alone tested it in a full-scale war.

AWACS, JSTARS, ABCS, ACCS: built for a different war

NATO's existing Command and Control (C2) systems — the AWACS airborne early warning aircraft, the JSTARS ground surveillance system, the Army Battle Command System (ABCS), the Air Command and Control System (ACCS) — were designed for a specific threat environment: the Cold War scenario of conventional Soviet ground forces advancing through the Fulda Gap into Western Europe. They are optimized for tracking massed armored formations, coordinating air interdiction, and managing large conventional force-on-force battles. They were not designed for a war dominated by thousands of drones operating simultaneously at low altitude in a contested electromagnetic environment.

The result is a doctrinal and technical gap between what NATO's existing systems do well and what the Ukraine battlefield demands. NATO systems struggle with drone swarm management, low-altitude air picture compilation, and the processing of the sheer volume of ISR data that modern drone warfare generates. Asgard was built specifically for this environment — which is why NATO planners are paying close attention.

Acquisition cycles vs. battlefield cycles

There is a structural tension between how NATO acquires military systems and how Ukraine develops them. NATO acquisition cycles, with their requirements definition, competitive procurement, development, testing, and fielding phases, typically span 5 to 15 years. Asgard was developed and deployed in years, not decades, driven by the immediate operational need of a fighting army. The gap between these timelines represents a genuine strategic risk for NATO: the next war may arrive before Western acquisition processes have completed the systems designed to fight it.

Ukraine's compressed development timeline was enabled by several factors: the existential urgency that focuses resources on immediate operational needs, a tech sector with deep software expertise (Kyiv was a major European software hub before 2022), and a military willing to field imperfect-but-useful systems rapidly rather than wait for perfect-but-unavailable ones. These are lessons that NATO's acquisition culture finds structurally difficult to internalize.

Five to fifteen years to develop and field a C2 system. That is NATO's timeline. Ukraine built Asgard and deployed it in a fraction of that time. I am not suggesting that NATO's processes are simply bureaucratic waste — they incorporate important safeguards, interoperability standards, and accountability mechanisms. But I am suggesting that the gap between these timelines represents a structural vulnerability that NATO's adversaries are aware of and will exploit.

AI on the armored vehicle: GYURZA-02

Asgard is part of a broader ecosystem of AI-assisted Ukrainian military systems. The GYURZA-02 armored personnel carrier, recently upgraded and entering service with Ukrainian forces, incorporates AI-powered battlefield vision — automated target detection, threat classification, and engagement recommendation for the vehicle crew. This brings the same kind of AI assistance that Asgard provides to operational planning down to the individual platform level: a crew in an armored vehicle benefits from machine vision that can detect, classify, and prioritize threats faster than human perception alone.

The GYURZA-02 upgrade represents the vertical integration of AI assistance throughout the Ukrainian force structure — from the corps-level planning that Asgard handles down to the individual platform engagement decisions that systems like GYURZA-02 support. This coherence of AI integration across multiple levels of command is a capability that most militaries have not yet achieved even in concept, let alone in deployed systems.

7,000 neutralized UAVs: the counter-drone network

Ukraine has also showcased a complete counter-drone network that, according to official figures, has neutralized more than 7,000 enemy UAVs. This network integrates multiple detection and engagement methods: radar, electro-optical sensors, electronic warfare jamming, dedicated interceptor drones, and small arms. The coordination of these diverse methods at scale requires exactly the kind of AI-assisted data fusion and response coordination that Asgard and related systems provide.

The ability to coordinate 7,000 drone neutralizations across a large operational theater, integrating data from dozens of sensor types and coordinating responses across multiple units, is a demonstration of AI-assisted battle management at operational scale — not in a test environment but in active combat against a sophisticated adversary using the full range of electronic countermeasures. This is Asgard's operational environment, and the 7,000 figure is one data point in its real-world performance record.

7,000 enemy drones neutralized. That number represents 7,000 individual detection-identification-engagement cycles, coordinated across a distributed network, under electronic warfare pressure. The human operators in that network could not have processed that volume manually at the required speed. AI assistance is not a luxury in this environment — it is the only way the math works.

Krasukha-4 and Borisoglebsk-2: Russia's electronic warfare arsenal

Asgard's dependence on data connectivity and drone-provided ISR creates a specific vulnerability: it can be degraded by effective electronic warfare (EW). Russia operates sophisticated EW systems including the Krasukha-4 — a mobile ground-based radar jamming and suppression system — and the Borisoglebsk-2, a tactical EW complex capable of jamming GPS signals, communications links, and drone control channels across a significant operational radius.

If Russian EW systems successfully deny GPS navigation and communications links to the drones that provide Asgard's ISR data, the system's informational picture degrades rapidly. Without real-time drone feeds, satellite imagery must substitute — which introduces delays. Without GPS-accurate targeting data, strike precision declines. Asgard is powerful in an electromagnetically permissive environment; its performance in a fully contested electromagnetic environment is a key open question.

Trust calibration: the human-AI decision boundary

A more subtle vulnerability is the trust calibration challenge inherent in any AI-assisted decision system. If commanders come to over-rely on Asgard's recommendations — treating its strike plans as authoritative rather than advisory — they risk losing the independent judgment that military command requires. An AI system trained on past data may also generate recommendations that are locally optimal based on historical patterns but miss emergent adversary behaviors that fall outside its training distribution.

Ukrainian commanders using Asgard are acutely aware of this risk, according to defence-ua.com's reporting on the system. The human-in-the-loop architecture is designed precisely to prevent automation bias — but maintaining genuine intellectual independence from AI recommendations, under the time pressure and cognitive load of active combat, requires ongoing cultural discipline in addition to technical design. This is a challenge every military deploying AI decision-support systems will face. Ukraine is figuring it out first.

Over-reliance on AI recommendations in military decision-making is a real risk. History offers endless examples of commanders who followed their models' recommendations into disaster. Asgard's designers built in the human override. But a piece of software cannot build the cultural habit of intellectual independence that prevents commanders from treating AI output as infallible. That is a training challenge. A leadership challenge. A human challenge that no algorithm solves.

From software hub to battlefield AI

Before the 2022 invasion, Kyiv was one of Europe's significant software engineering centers — home to the Eastern European offices of major Western tech companies, numerous successful software service firms, and a graduate engineering pipeline from Ukraine's strong technical universities. That human capital did not disappear when Russia invaded. Much of it was redirected — into drone production, into battlefield software, into AI systems like Asgard and DELTA.

This software-to-warfare technology transfer is one of the most remarkable human resource stories of the war. Engineers who were writing enterprise software in 2021 are writing targeting algorithms in 2026. Developers who were building fintech applications are now working on electronic warfare countermeasures. Ukraine's wartime defense tech industry is built on a civilian software foundation that most military planners had not considered a strategic asset — until it proved to be one.

The ARX Industries model: German-Ukrainian robot production

The Asgard story fits into a broader pattern of Ukrainian defense technology development that is now attracting international partnerships. ARX Industries, a German-Ukrainian joint venture, is building thousands of frontline robots for deployment on Ukrainian positions. TAF Industries signed a drone manufacturing deal with Polish state defense group PGZ at the Gdańsk conference. These partnerships bring Ukrainian combat experience and software innovation together with Western industrial capacity and capital.

Asgard's next evolution may follow the same path — Ukrainian AI architecture and combat-tested training data combining with NATO member industrial partners to create systems deployable across the Alliance. Whether that path leads through formal procurement or through the kind of rapid field adoption that characterizes Ukraine's development model remains to be seen. But the direction is established: Ukraine's battlefield AI is moving from domestic innovation to allied adoption.

Engineers who wrote fintech apps in 2021 are writing targeting algorithms in 2026. That sentence contains an entire industrial mobilization story. Ukraine did not have a large pre-war defense industry. It had a large tech sector and a population under existential threat. It converted one into the other with a speed that no industrial policy, no defense procurement reform, no talent development program could have engineered deliberately. Necessity made it happen.

The global debate Ukraine is forcing

Asgard's development is occurring in the context of a global debate about the ethical limits of AI in warfare. Companies like Axon — the US body camera and drone manufacturer — have invested in drone-AI integration specifically for law enforcement applications, and those technologies have inevitable dual-use implications. The Ukrainian experience is forcing the pace of this debate: while Western democracies discuss AI ethics frameworks at conferences, Ukraine is deploying AI battle management systems in real combat and deriving operational lessons in real time.

The "fourth law of drone warfare" emerging from this experience — after the laws of range, payload, and cost — might be stated as: the military that automates decision support fastest, while maintaining meaningful human control, will have a decisive tempo advantage over one that moves slower to preserve familiar processes. Ukraine, constrained by necessity, is writing that law from operational experience. NATO members observing from the outside need to decide whether to learn from it or wait to rediscover it in a future conflict of their own.

What responsible AI in warfare looks like

Asgard's architecture — AI-generated recommendations, human authorization required for execution, continuous learning from combat outcomes — is one practical implementation of responsible AI in military applications. It is not the only possible implementation, and it will not be the last word in a rapidly evolving field. But it demonstrates that meaningful human control and AI-accelerated decision tempo are not mutually exclusive — they can be designed together, if the architecture is built with human oversight as a first-order requirement rather than an afterthought.

Ukraine's development of Asgard under the most extreme operational pressure imaginable, while preserving the ethical principle of human decision authority, is a contribution to the global conversation about military AI that deserves more recognition than it has received. The country on the front lines of drone warfare is also on the front lines of thinking about how to fight with AI in a way that keeps humans responsible for the consequences.

Meaningful human control and AI tempo acceleration, designed together. That is what Asgard claims to offer. I am not in a position to audit the system's code or verify every claim about its performance. But I note that the people building it are doing so under conditions where a wrong decision — whether it comes from a flawed algorithm or a distracted human — gets people killed. That operational accountability concentrates the mind in ways that no ethics review board can replicate.

The math of daily strike intensity

If Asgard genuinely enables 10 times more daily strikes than the planning processes it replaces, the operational implications are profound. Currently, Ukrainian forces conduct a certain number of targeted strikes per day across the front — strikes on command posts, logistics nodes, artillery positions, air defense systems. Multiply that number by 10, with comparable targeting accuracy and strike effectiveness, and the rate at which Russia's military infrastructure is degraded changes dramatically.

More strikes per day means more pressure on Russian logistics, more Russian command posts disrupted, more artillery systems destroyed, more Russian operational planning cycles interrupted. The compounding effect over weeks and months would be a Russian force that cannot reconstitute and regenerate as fast as Ukraine can degrade it — which is the definition of attrition advantage in a war of this character.

Verification and humility: what we cannot know yet

The 10x claim and the 60-minute planning cycle, for all their compelling logic, must be evaluated with appropriate skepticism. System performance claims made by developers of military software are not peer-reviewed publications. The operational conditions that produce a 60-minute cycle under one set of circumstances may not generalize to other conditions. The 10x strike multiplier assumes that planning is the binding constraint on strike rate — which may not always be true when the binding constraints are weapons availability, target generation, or risk tolerance.

None of this means the claims are false. It means they should be understood as performance targets in optimal conditions, to be refined by operational experience, rather than as guaranteed minimum performance specifications. Ukraine's willingness to deploy and learn from imperfect systems rapidly is itself a capability — but it means the claims associated with those systems should be read with that context in mind.

I am enthusiastic about Asgard. I am also committed to intellectual honesty. The 60-minute cycle and 10x multiplier are claims that deserve testing, scrutiny, and operational validation over time. Military technology journalism has a history of overhyping capabilities that turned out to be more modest than advertised. I note that risk while genuinely believing that what Asgard represents — AI-assisted battle management developed under real combat conditions — is one of the most significant military innovations of this war.

What Ukraine has built and what it means

Asgard — a battle management AI trained on four years of real combat data, integrated with the full sensor network of Ukraine's military, and capable of compressing corps-level planning cycles from 72 hours to 60 minutes — is a demonstration of what is possible when necessity, engineering talent, and battlefield urgency converge. Whether every performance claim proves out under scrutiny or not, the system represents a genuine breakthrough in the pace at which military planning can operate at scale.

The implications extend far beyond Ukraine. Every military that faces or anticipates high-intensity conflict is watching Asgard's development — whether they acknowledge it publicly or not. NATO's C2 systems were built for a war that Ukraine's experience has shown is already obsolete. The planning cycle that takes 72 hours in conventional doctrine needs to be measured in hours — or minutes — in the drone-dense, electromagnetically contested warfare of the twenty-first century. Ukraine is figuring that out first. The rest of the world should be paying tuition.

The honor of being a student of necessity

There is a peculiar honor in what Ukraine has achieved with Asgard. This is not a system developed in comfort, with generous budgets and unlimited time. It was developed under fire, by engineers and commanders who knew that the difference between a working system and a failed one might be measured in lives. That pressure produced something remarkable: a real-world demonstration that AI-assisted military planning at machine speed is not a future concept — it is a present capability.

Ukraine built it because it had to. NATO should adopt it because it has the chance to learn before having to. The 60 minutes that Asgard buys might one day be the difference between a successful defense and a catastrophic failure — not just in Ukraine, but in any future theater where speed of decision defines survival.

By Maxime Marquette, columnist

Editorial stance

I am not a journalist but a columnist and analyst. My expertise lies in observing and analysing the geopolitical, economic and strategic dynamics that shape our world. My work is to take political strategies apart, to understand global economic movements, to put the decisions of international actors in context and to offer analytical perspectives on the changes that are redefining our societies.

I do not claim the cold objectivity of traditional journalism, which stops at reporting facts. I claim analytical clarity, rigorous interpretation and a deep understanding of the complex issues that concern us all. My role is to give meaning to the facts, to place them in their historical and strategic context, and to offer a critical reading of events.

Method and sources

This text keeps the basic distinction between verified facts and interpretive analysis. The rule never changes: a factual claim is published only when a verifiable source supports it, and the sources this article actually used are listed under Sources, never here.

Primary source families used by the publication, when the story has them: official statements from governments and international institutions, public statements by political leaders, reports from intergovernmental organisations, dispatches from recognised international news agencies.

Secondary source families: specialised publications, internationally recognised news media, analyses from established research institutions, reports from industry organisations.

When an article cites statistical, economic or geopolitical data, it comes from the institutions that produce it (intergovernmental organisations, central banks, national statistics offices), and the exact institution is named under Sources.

Nature of the analysis

The analyses, interpretations and perspectives in the analytical sections of this article are a critical, contextual synthesis based on the information available, the trends observed and the expert comments cited in the sources consulted.

My role is to interpret these facts, to set them in the context of today's geopolitical and economic dynamics, and to give them a coherent meaning within the larger story of the changes shaping our time. These analyses reflect expertise built through continuous observation of international affairs and an understanding of the strategic mechanisms that drive global actors.

This article describes a documented state of affairs on its publication date, not a prediction: later developments may shift its outlook. No update is promised in advance; when an article is corrected or completed, the change is dated in the text.