Subscribe to The Informer for monthly expert analysis, and to Events for advance notice of visiting world leaders and distinguished guests.
You may unsubscribe from Lowy Institute newsletters at any time. For information on our privacy practices and how to unsubscribe, see our Privacy Policy.
Subscribe to The Informer for monthly expert analysis, and to Events for advance notice of visiting world leaders and distinguished guests.
You may unsubscribe from Lowy Institute newsletters at any time. For information on our privacy practices and how to unsubscribe, see our Privacy Policy.
Battlefield evidence is everywhere. The will to act on it is the missing Western capability.
Modern war and the systemic learning deficit in Western military institutions
About the author
Mick Ryan
Mick Ryan is a Senior Fellow for Military Studies in the Lowy Institute’s International Security Program.
This paper identifies a critical strategic vulnerability: Western military institutions, including in Australia, are failing to energetically learn from modern wars. Despite four years of unprecedented visibility into Ukrainian battlefield innovations, and the recent war in Iran, Western forces have not institutionalised key lessons into doctrine, force structure, or procurement priorities.
This has implications for a range of strategic competitions. China, Russia, Iran, and North Korea have built an authoritarian knowledge market in which battlefield insights transfer rapidly. Western democracies, constrained by risk-averse bureaucracies, slow procurement cycles, and promotion systems rewarding managerial competence over innovative thinking, face a structural disadvantage in this competitive learning environment.
For Australia, the consequences are serious: minimal drone capabilities, almost no counter-drone defences for deployed forces or critical infrastructure, the opportunity costs of expensive systems, and slow mechanisms for translating foreign war lessons into force development.
The paper’s five recommendations address culture change, promotion reform, AI-enabled learning, rapid drone capability development, and acquisition reform. These are essential adaptations. The alternative is an Australian Defence Force structurally unprepared for Indo-Pacific contingencies where the lessons of the Ukraine war — massed autonomous systems, drone-enabled combined arms, and indigenous rapid production — are highly likely to be directly applicable. Ignoring visible evidence from modern wars when adversaries are absorbing those same lessons at speed constitutes a strategic choice with potentially catastrophic consequences for military organisations — and entire nations.
Technological advances will not change the essential nature of war. Fighting will never be an antiseptic engineering exercise. It will always be a bloody business subject to chance and uncertainty in which the will of one nation will be pitted against another, and the winner will be the one that can inflict more punishment and absorb more punishment than the other side. But the way punishment gets inflicted has been changing for centuries, and it will continue to change in strange and unpredictable ways. — Max Boot
In 2023, a Russian A-50 airborne early warning aircraft was attacked on the ground at the Machulishchy airbase near Minsk. Located 200 kilometres from the Ukrainian frontline, the Russian Air Force had not imagined, nor prepared for, an attack on this location. In March 2026, a United States Air Force (USAF) E-3 Sentry airborne early warning aircraft, parked in the open at a Saudi Arabian airbase approximately 700 kilometres from Iran, was destroyed during Operation Epic Fury, the American–Israeli campaign against Iran. There was one key difference between these attacks: the USAF had years of warning about the threat, which it did not heed. It demonstrated a lack of learning from other people’s wars.
At the strategic level, an even more recent glaring failure to learn is obvious. The Trump administration has failed to learn the central political lesson from the war in Ukraine: even supposedly much weaker nations in a war have agency. Such belligerents can demonstrate the will to resist foreign military aggression for years, if needs be. This has been the case for over four years in Ukraine and appears to be the case in the Iran war.
The contrast between Western institutional learning inertia and the speed of adversarial learning is one of the defining strategic facts of this decade. Western governments and militaries have been slow to institutionalise the lessons of Ukraine and Iran. Their adversaries have not. China, Russia, Iran, and North Korea have built an authoritarian knowledge market in which battlefield insights — from drone employment to electronic warfare, from industrial mobilisation to strategic coercion — flow more rapidly than many Western institutions have acknowledged. When one member of this bloc learns, all of them can learn.
Organisations can exhibit persistent failures to learn. The foundational work of theorists Chris Argyris and Donald Schön distinguished between “single-loop” learning — correcting errors within existing frameworks — and “double-loop” learning, which requires questioning the underlying assumptions themselves. Military institutions are archetypal single-loop learners. They excel at tactical adaptation within doctrinal boundaries but can be structurally resistant to revising core doctrines. Organisations can fall into “competency traps” where they reinforce familiar routines even when those routines no longer serve the environment.
While military units in the West have demonstrated an admirable vigour to learn the lessons of foreign wars, this same energy has not been apparent in their broader military and political institutions. Indeed, mainly because of the speed of change, much of the transformation in war of the past half-decade appears to have eluded defence bureaucracies. This deficit of learning is due to a combination of emphasis on pre-existing ideas, failure to accurately understand what occurs in foreign wars, application of disputed or misleading lessons, and failure to sustain the implementation of useful lessons.
A learning deficit afflicts Western military institutions, including the Australian Defence Force (ADF). This is not a resource problem nor a deficit of relevant information. It is a challenge of organisational culture, individual and institutional humility, leadership philosophies, and political inattention. In an era when war has never been more visible, choosing not to learn is a strategic decision with grave consequences.
This paper contains four sections. The first examines the nature of military learning and why it is so difficult to do well in peacetime. The second and third sections offer case studies in Western learning failure: the counter-drone war and the evolution of offensive operations. The fourth section draws together the analysis into recommendations for the Australian Government and the Department of Defence. The paper concludes with a call for a different kind of institutional leadership, one prepared to nurture “responsible rebellion”.
History offers a rich and often cautionary literature on how military institutions learn, how they fail, and how the consequences of those failures are measured. The core findings of this literature provide the analytical foundation for what follows.
Cognitive science identifies schema rigidity — the tendency to interpret new information through pre-existing frameworks rather than updating those frameworks — as a core mechanism in both individual learning disabilities and organisational inertia. In military organisations, doctrine functions as a collective schema. Lessons from another country’s conflict that confirm existing doctrine are readily absorbed; those that challenge it are marginalised, reinterpreted, or ignored.
Adaptation theory, drawing on Darwin’s foundational insights about evolutionary pressure and natural selection, provides a useful construct for identifying remedies to military learning disorders. A core insight from the literature on this topic is that institutions that can identify change in their environment, analyse its implications, and alter their behaviour accordingly will outcompete those that cannot, or do not. Military organisations exist in a permanently competitive learning environment, in peace and in war, even if they do not always fully appreciate the degree of change and competition they face. This drives the need to build and continuously evolve learning and adaptation cultures; cultures that incorporate both individual competence and institutional processes.
Adaptation in military organisations does not occur at a single moment or in a single mode. Three distinct forms of adaptation are relevant to the contemporary challenge: adaptation before war, adaptation during the transition from peace to war, and adaptation in war. Each has different organisational and leadership imperatives, each proceeds at a different pace, and each requires a different institutional disposition.
Peacetime adaptation is the foundation on which wartime performance rests. An institution’s proficiency with organisational learning in peacetime will have a significant impact on how it prepares for conflict, how it performs in the early stages of war, and how quickly it can close the gap between its pre-war assumptions and wartime reality. As Williamson Murray, one of the foremost scholars of military effectiveness, has argued, military culture may be the most important factor not only in military effectiveness but also in the processes involved in military innovation, which is essential to preparing military organisations for the next war.
Peacetime adaptation, however, demands that institutions navigate a permanent tension between exploitation and exploration. Exploitation means applying existing competencies to well-understood problems. Exploration means investigating new competencies and solutions to emerging challenges. These two are always in tension. As organisational theorist James March writes, “adaptation requires a balance of exploration and exploitation but is continually threatened by the tendency of each to extinguish the other”. Western military organisations, operating in peacetime and under the fiscal and bureaucratic pressures of liberal democratic governance, have in recent years tended sharply towards exploitation. They are doing what they have always done, with incremental refinement, at the expense of genuine exploration.
Learning lessons from historical or ongoing wars is a complicated endeavour. Even concerted and well-resourced efforts to collect and implement lessons are no guarantee of success. The military learning literature identifies several persistent themes about the challenges to effective institutional learning.
The first is organisational culture — the beliefs, ideas, norms, and assumptions embedded in an organisation, often below the level of conscious articulation — which exerts a powerful influence on an institution’s capacity to learn and change. It shapes the degree to which leaders at every level recognise that change is occurring and how much risk they will accept in response. A culture that rewards conformity, penalises failure, and promotes leaders on the basis of managerial competence rather than innovative thinking will produce a poor learning organisation, regardless of the formal processes it has established.
The second obstacle is the paradox at the heart of lesson-learning: highly specific lessons provide little value for applying to new cases, while general lessons that might be applicable across cases can be so banal as to offer no insight. This paradox is amplified when the learning institution differs significantly from the institution being observed — different geography, different force structure, different strategic context. This is relevant when proposing lessons from Ukraine for application in the Pacific or Middle East.
Some Western military institutions appear to have reached the conclusion that Ukraine’s experience is so specific that its lessons cannot be straightforwardly applied to a force preparing for other contingencies, such as in the Pacific. A clear example was the 2024 Australian National Defence Strategy, which largely ignored the war. This was partially redressed in the 2026 version. It is true, as US military strategist Joseph Collins notes, “technology-inspired lessons from a single war are likely to have a very short life”. But this also fosters a system of learned helplessness when it is used as a reason to avoid the hard intellectual endeavour of determining which lessons can transfer.
A third obstacle is the tendency towards rapidly drawing preliminary findings and seeking hard data to support them. This orientation discourages attention to operational and strategic elements and promotes static, not dynamic, learning. In other words, institutions look for what they already expect to find and stop looking once they have found it. Examples include the British and their observations from the Russo–Japanese War, the US Army’s observations of the Yom Kippur War, and multiple observers of the Nagorno–Karabakh War.
A fourth obstacle, and perhaps the most fundamental, is the humility deficit: a failure to maintain genuine intellectual openness about what an adversary or a war might reveal. There are three prominent failures of military humility in recent years: Russia’s catastrophic underestimation of Ukrainian national resilience in February 2022; the collective failure by Ukraine and its Western supporters to understand the depth and sophistication of Russian defensive preparation before the 2023 counteroffensive; and Israel’s failure to appreciate Hamas’s operational planning before 7 October 2023. In each case, the failing institution possessed substantial intelligence. What it lacked was the willingness to take that intelligence seriously, to follow its implications, and to adapt before being surprised.
The most disorienting feature of adaptation to modern war is its pace. Ukrainian drone units often adapt and update their software daily. Tactics at the unit level evolve every one to two weeks. Russian and Ukrainian combined-arms tactics at the higher level evolve on a cycle of two to three months. Against this tempo of change, Western military institutions, operating on procurement cycles measured in years, doctrine revision cycles measured in decades, and leadership selection processes that consistently reward managerial over innovative competence, are structurally misaligned. Military scholar Theo Farrell has written that “when armies, air forces and navies go to war, invariably they do so with plans, tactics and equipment not optimised for the fight ahead. Hence the ability to adapt quickly in war can be decisive.” Learning quickly matters. This can be assisted by pre-war development of systemic learning and adaptation functions in an institution.
The nations involved in the wars examined in this paper have begun to build, in live operational conditions, AI-enabled learning infrastructure that can speed up learning and adaptation. Artificial intelligence (AI) in military organisations tends to be associated with targeting and planning. AI-enabled target identification, weapons allocation, and kill-chain acceleration has attracted the bulk of public attention and generates the most significant ethical debate. However, AI is starting to be employed for learning and adaptation. AI applied to the aggregation of battlefield data and rapid translation of operational experience into institutional knowledge is less visible but critical to the learning failures diagnosed here.
Ukraine has built what is now one of the world’s most consequential live battlefield AI learning platforms. The OCHI system, a Ukrainian non-profit initiative, centralises video feeds from more than 15,000 frontline drone crews. Since 2022, it has aggregated two million hours of drone footage — equivalent to approximately 228 years of continuous observation — from which AI models are trained on combat tactics, target recognition, and assessing the effectiveness of weapons systems. This is not AI for targeting. It is AI applied to the problem of learning: building a continuously updated model of what works, what fails, and why, from the largest dataset of contemporary combat ever assembled. In March 2026, Ukraine opened this infrastructure to multiple AI vendors simultaneously, establishing the first multi-party platform for training AI models on near-real-time battlefield data, specifically designed to close the gap between laboratory simulation and unpredictable operational conditions. The practical consequences of this investment are already apparent: the integration of AI guidance into Ukrainian first-person view (FPV) drone operations has increased strike effectiveness from 30–50 per cent to 80 per cent.
The American data analytics company Palantir’s deployment in Ukraine provides a second model. Partnering with the Pentagon in Project Maven, Palantir built the Maven Smart System which fuses intelligence from drones, satellites, ground sources, radar, and thermal imaging, presenting commanders with AI-generated targeting options that improve with each engagement. The critical detail is that during the first ten months of Maven’s operational support to Ukraine, the system underwent more than 50 rounds of improvement. This is AI adaptation at a rate that has no parallel in the Western doctrine revision or procurement cycles. It represents a qualitatively different institutional relationship with learning in which the feedback loop between operational experience and system improvement is measured in days.
Israel has institutionalised AI for military learning through dedicated organisational architecture rather than through a single platform. Unit 8200 of the Israel Defense Forces (IDF) maintains an embedded AI Centre whose mandate includes continuous system improvement from operational data; Unit 3060, within the Intelligence Division, was established specifically to advance operational and visual AI systems for ongoing combat use. The pace of the IDF’s AI adoption accelerated sharply after the 7 October intelligence failure. The IDF rapidly cleared AI tools for deployment following the attack, with technology company reservists providing capabilities not previously available within the military establishment. This is the inverse of the pattern this paper diagnoses in Western institutions: crisis produced acceleration rather than inertia.
In the US military, most of the AI work that occurred in 2025 remained concentrated in four functional lanes: decision support, intelligence processing, maintenance, and training. The Marine Corps formalised its service-level AI plan only in September 2025. The Pentagon’s AI Rapid Capabilities Cell was launched in December 2024. US combatant commands have also employed AI for a range of functions including wargaming and operational learning.
Ukraine, Israel, and the United States have demonstrated at speed and at scale, in live operational conditions, that AI can function as a learning accelerant. It can act as a tool for aggregating battlefield experience, identifying patterns across thousands of engagements, and translating operational data into institutional knowledge at a pace no human analytical process can match. Many Western military institutions, including the ADF, have watched this demonstration. They have not replicated it. The case studies that follow illustrate what that failure might cost.
Building effective learning cultures requires the right conditions across several reinforcing dimensions: contextual appreciation of the environment in which lessons are generated; an organisational mindset that accepts risk, and rewards innovation; formal processes for collecting, analysing, and disseminating lessons; and leadership that champions adaptation over conformity.
In particular, new promotion pathways are a crucial element in unearthing adaptive leaders and allowing them to lead in less conventional ways that nurture and champion innovation. Political scientist Barry Posen’s The Sources of Military Doctrine argues that intervention by civil authorities in military reform is often necessary to enable doctrinal change because peacetime promotion systems tend to entrench existing paradigms. Williamson Murray and Allan Millett, in Military Innovation in the Interwar Period, showed that across multiple cases such as carrier aviation, armoured warfare, and amphibious doctrine, the ability to promote and protect innovative officers was decisive in determining whether theoretical advances translated into operational reality. A promotion system that rewards those who challenge conventional wisdom, invest in doctrinal innovation, and demonstrate adaptive leadership in novel conditions will produce an officer corps capable of transformation.
The historical record demonstrates that military organisations can learn: the Prussian general staff reforms, the interwar development of carrier aviation and armoured warfare, the British Army’s transformation under fire in the First World War, and Ukraine’s own wartime adaptation since 2022 all stand as evidence. But success has never been automatic; it has required institutions that made structural space for change. The case studies that follow illustrate what happens when that space is absent.
The war in Ukraine since February 2022 has generated a robust body of counter-drone operational learning. Russia’s escalating campaign of mass drone strikes, which now average between 3,000 and 5,000 long-range attack drones per month against Ukraine, has created an existential counter-drone requirement that Ukraine was forced to solve under fire. Every phase of Ukraine’s response has been covered in open sources, documented by think tanks, and briefed to allied governments.
Against this backdrop, defence scholar Michael Horowitz has observed a fundamental shift in modern warfare. He has proposed the collapse of the binary between mass and precision; between scale and sophistication. This shift is most visible in the counter-drone domain. Ukraine’s answer to the threat was cheap, mass-produced, rapidly adaptable interceptor drones, a solution that inverted the economics of air defence. Western military institutions, watching this unfold, have continued to invest primarily in expensive, low-density counter-drone systems. The information was always available; the limiting factor was institutional willingness to act on it.
Ukraine’s response to the growing Russian drone threat was a masterclass in structured learning under fire, a continuous cycle in which each Russian tactical innovation — new drone routes, new frequencies, new electronic countermeasures, thermobaric warheads on Shahed and Geran drones — has been met by a corresponding Ukrainian adaptation in detection systems, interception technologies, and employment doctrine.
The most significant Ukrainian counter-drone innovation, and the most glaring instance of Western failure to implement what was demonstrated in plain view, is the development of low-cost drone interceptors. By late 2024, Ukraine had developed and deployed interceptors costing $1–5,000 that could reliably defeat approximately 70 per cent of incoming Russian Shahed drones, which cost ten times as much. The economics of counter-drone warfare had been inverted. A cheap, locally manufactured, rapidly adaptable solution had fundamentally changed the cost calculus of strategic drone strikes.
Several nations entered partnerships with Ukraine to co-produce cheap drones and interceptors, with manufacturing facilities established in Ukraine and across Europe. By early 2026, Ukraine had manufactured 100,000 interceptor drones, with production growing. The average daily supply of FPV interceptor drones has now reached 1,500, each of which is integrated with radar systems to enable automated or semi-automated launch and mobile fire groups against massive nightly attacks by Shahed-type loitering munitions.
The economics of this capability are central to understanding the nature and depth of Western institutional failure. A single Patriot interceptor costs over $3 million; a NASAMS round slightly over $1 million. Each Shahed costs Russia as little as $35,000 to manufacture. Ukrainian drone interceptors, by contrast, cost between $1,000 and $5,000 apiece, with an average interception rate exceeding 90 per cent by March 2026. This cost-exchange problem represents the defining challenge for Western air defence planning in the coming decade, requiring a fundamental reorientation away from anachronistic contracting and acquisition systems, and the structural bias towards expensive systems that dominate Western procurement culture. That reorientation has not happened. The limiting factor has not been lack of knowledge of the threat or of the solution. It has been a lack of institutional willingness to make the change.
This failure is not confined to any single nation. It has been a collective Western institutional failure across the United States, NATO, Europe, and Australia. A 2024 investigation by Defense One found that at some of the US military’s key centres for studying warfare, the services appeared to treat the grinding but tech-heavy war in Ukraine as just one topic among many. The Marine Corps had not directly interviewed Ukrainian military personnel, relying instead on other services and allies to gather first-hand information. The US Navy was unable to describe its approach to Ukraine war lessons at all.
The US Congress, at least, was paying attention: the Senate National Defense Authorization Act for Fiscal Year 2025 required the military to regularly report to Congress on efforts to identify, disseminate, and implement knowledge gained from the war. That such a requirement needed to be legislated into existence is itself a measure of the depth of the institutional failure. Some progress was made in 2026, with the US Army announcing a force-wide effort to redevelop all doctrine for drone usage. But that announcement came more than four years after the 2022 full-scale invasion, and nearly 18 months after Ukraine had first demonstrated the viability of cheap interceptors at operational scale.
NATO’s counter-drone posture has been equally slow to adapt. There remain significant gaps in alliance-wide doctrine and procurement. European nations that did acquire some counter-drone capability frequently lacked appropriate doctrine, training pipelines, procurement policies and operational concepts to employ it effectively. NATO has belatedly begun to address this threat in the wake of several incidents in 2025 when Russian drones penetrated NATO airspace with its Layered Counter-UAS Initiative (LCI-X).
The technology transfer failure is perhaps the most damning dimension of this story. Ukraine consistently and openly offered to share its counter-drone experience. The Ukrainian ambassador to Australia noted at a 2025 defence conference that his country possessed millions of hours of battlefield video and operational data and expressed willingness to share it with allied governments through government-to-government arrangements. The limiting factor was not knowledge. It was institutional willingness to receive and implement it.
Ukraine first demonstrated cheap drone interceptors at operational scale in late 2024. The first reported Pentagon procurement inquiry into Ukrainian interceptor technology came in March 2026, a gap of more than 18 months during which Western institutions observed, were briefed, and did not act.
The Iran conflict of early 2026 transformed the counter-drone learning failure from a professional embarrassment into a strategic liability. When the United States and Israel commenced operations against Iran, the US Secretary of Defense was forced to brief Congress that American air defences could not intercept all of Iran’s Shahed-class drones. The United States, with the wealthiest and most technologically sophisticated military in the world, was then reported to have asked Ukraine for assistance in countering the very drone threat that Ukraine had been managing, and offering to help allies manage, for more than two years.
The contrast could not have been starker. The drones Iran employed were of the same family — the Shahed — that Ukraine had been intercepting at scale for more than a year. Analysts had long observed that Western planners assumed technological superiority and sophisticated air defences would ensure control of the skies; mass-produced drone proliferation had challenged that assumption. Ukraine had demonstrated that the answer to cheap, mass-produced drone attacks was not better missiles but cheap, mass-produced interceptors designed to complement traditional air defences. The United States had watched this demonstration for two years and had not replicated it. The information was always available.
For Australia, the consequences of this collective Western learning failure are acute. The ADF currently deploys a relatively small number of military drones and has no armed drones at operational scale. Major armed drone programs have been cut by successive governments and cost-cutting imposed on the Army has reduced funding for innovation. The ability of the ADF to conduct counter-drone operations for deployed forces and critical infrastructure is very limited.
This reflects a pattern of procurement priorities that remains addicted to small quantities of expensive platforms built overseas — such as the Triton unmanned maritime surveillance aircraft — rather than the cheap, mass-produced systems that Ukraine’s experience has demonstrated are operationally decisive. The Asia Group found in 2025 that Australia has not yet established a dedicated regulatory framework for C-UAS. Unlike the United States, Australia lacks even a legislative mechanism requiring the military to report on its implementation of Ukraine war lessons. Inflation has compounded the problem, decreasing the purchasing power of Australia’s defence dollar by more than seven per cent since 2020. While the 2026 National Defence Strategy indicates some movement, the deeper issue is conceptual: defence planners have continued to see uncrewed systems largely through the lens of a pre-Ukraine mental model.
The Ukrainian ambassador’s warning at the Indo Pacific 2025 International Maritime Exposition is worth highlighting. He told his Australian audience that technologies which were state of the art three years ago are now obsolete. This was made so by Russian electronic warfare and countermeasure development, and yet allied nations did not know this because they had not tested their own systems against contemporary countermeasures. He warned explicitly that Australia’s Bushmaster vehicles are now vulnerable to drone attack in ways that would have been difficult to anticipate before 2022. This was not a warning about offensive drone capability; it was a warning about the counter-drone gap. This gap is defined by the failure to develop, procure, and field the means to protect platforms and personnel from a threat that has been visible and documented for more than four years.
The organisational consequences of this conceptual failure are visible across the ADF. There are few dedicated drone units and no dedicated counter-drone units at operational scale. There is no independent unmanned systems force to act as an institutional advocate and no publicly documented doctrine for counter-drone employment at the tactical and operational levels. Reserve components might be better organised to contribute to the counter-drone mission. Operation Kudu — Australia’s contribution to the multinational training mission for Ukrainian recruits in the United Kingdom — has had an unexpected effect: Australian soldiers are learning about drone operations and other aspects of modern war from the very recruits they are training. Yet there is no systematic process for institutionalising those insights into doctrine, force structure, or procurement priorities.
Drone and counter-drone operations are the most visible dimension of the adaptation battle in Ukraine. But this is not the only dimension. The war has also produced an ongoing transformation in the conduct of offensive operations. This includes tactical ground operations and long-range strike. Here too, the gaps between what has been demonstrated in Ukraine and Iran, and what has been absorbed by Western military institutions, are consequential.
Early commentary on the Ukraine war was dominated by claims of defensive ascendancy — the return of the trench, and the apparent obsolescence of the armoured offensive. The scholarly debate on this question remains open. But the evidence presents a more varied picture, with successful offensive and defensive operations both apparent. Offensive innovation is occurring at the tactical level, and it has significant implications for any military that will fight a peer adversary.
Ukraine’s ground forces, operating under substantial manpower constraints and fighting a Russian military that has improved its own learning and adaptation, have nonetheless developed new offensive methodologies that represent an important conceptual innovation. The Ukrainian advance into Russia’s Kursk oblast in August 2024 demonstrated that tactical and operational manoeuvre, and offensive operations, remained possible even with the massed use of drones by each side. Tactical success may not always lead to strategic success, but offensive operations are still possible in the new drone era.
The most recent example of offensive manoeuvre has been described by Dmytro Filatov, commander of the 1st Assault Regiment, in open-source reporting from early 2026. He details a two-battalion attack that achieved a ten-to-twelve-kilometre penetration of Russian lines while subsidiary units flooded the grey zone, making diversionary attacks, destroying reconnaissance assets, encircling Russian outposts, and mopping them up.
The key innovation was the massed employment of drones as a decisive element of the combined-arms team. The first step of each attack was a saturation drone strike: a wave of 200–300 drones swarming over the selected assault area to hunt targets before ground troops moved. Once the ground troops followed, defending forces faced an impossible dilemma — remain in position and be found and destroyed by artillery or large, multi-munition-carrying bomber drones, or move and be rapidly detected and hunted by smaller, nimbler FPV drones. The first drone wave served simultaneously as a direct fire weapon, a reconnaissance tool, and a suppression mechanism. Concurrently, counter-drone teams defended advancing forces and denied Russian drone operations. This has changed the economics of C-UAS operations and offers a means for more assured offensive manoeuvre.
This is not a niche innovation. It represents a fundamental change in how combined arms operations are conducted at the tactical level. This is a development in the intellectual realm of the military profession that sees new technology being absorbed effectively and demands that drone units be as deeply integrated into manoeuvre planning as artillery, infantry, engineers, or armour. No Western army, including the ADF, currently possesses this capacity. Few are systematically developing it. Too many are still conceptualising drones as sensors or niche strike assets rather than as the decisive element of a combined-arms team.
Understanding offensive innovation in Ukraine also requires understanding how Russia has evolved its offensive approach. This is because any future adversary of Australia or its allies is more likely to resemble the adapted Russian military of 2026 than the rigid, centralised force that failed so visibly in 2022.
Jack Watling of the Royal United Services Institute has described the evolution of Russian offensive doctrine into what he calls the Russian Offensive Triangle. Russian ground forces first seek to fix Ukrainian frontline units with infantry and mechanised troops. They then limit Ukrainian manoeuvre and inflict attrition through FPV drones, loitering munitions, and indirect fires such as artillery and mortars but also scatterable mines placed between and behind Ukrainian positions. Finally, Russian air forces employ UMPK glide bombs against Ukrainian forces now fixed in their defensive positions by the preceding effects.
This approach poses a dilemma for a defender, forcing it to choose between holding static defensive positions, which reduces drone casualties but invites destruction by glide bombs, and retaining mobility, which avoids the glide bombs but increases vulnerability to drones. This dilemma is not specific to Ukrainian geography. Any defending force confronting a state adversary that has absorbed the Russian Offensive Triangle will face a version of it. ADF planners should be thinking hard about how Australian doctrine, force structure, and materiel choices address this challenge.
The second aspect of offensive adaptation in Ukraine is long-range strike. Ukraine’s development of increasingly sophisticated strike capabilities since 2022 — from modified Soviet surveillance drones to a family of indigenously produced cruise missiles with ranges up to 2,000 kilometres — represents an extraordinary case study in rapid defence-industrial innovation.
The pace of this development is striking. In August 2024, Ukraine announced the Palianytsia rocket-drone with a range of 500–700 kilometres. In December 2024, it revealed the Ruta long-range cruise missile with a range of 500–700 kilometres. In the past year, it has also unveiled the FP5 Flamingo missile (3,000 kilometre range), a longer-range variant of the Neptune missile (1,000 kilometre range), and the FP7 ballistic missile (an alternative to the US ATACMS system, with a 300 kilometre range). This generation of advanced precision strike systems, equivalent to a decade or more of normal Western procurement, was developed in less than three years.
Australia’s 2024 and 2026 National Defence Strategy documents identified long-range strike as a priority, allocating eight per cent of capability investment to targeting and long-range strike over the decade to 2034. The current plan skews heavily towards high-end, sophisticated, low-density systems. The evidence from modern conflict instead points to a mix that includes some large, exquisite systems, mass quantities of cheaper autonomous aerial, maritime, and ground systems, and investment in indigenous innovation and production. Ukraine’s experience provides the empirical foundation for this recommendation: indigenous systems that can be rapidly updated have proven operationally decisive and provide supply-chain resilience in a way that dependence on imported precision weapons cannot.
The learning deficit described in this paper is serious and structural. It will not be remedied by a white paper, a committee, or an additional ministerial statement. The essential task is a transformation of institutional culture — the beliefs, habits, leadership philosophies, and accountability structures that determine how the ADF and the broader national security enterprise actually learn, and how quickly they act on what they learn.
The recommendations that follow incorporate a range of government and military endeavours. They include proposals about cultural and leadership change; organisational reform; capability priorities; and alliance and partnership mechanisms. They are intended to be complementary and mutually reinforcing. Some can be implemented immediately. Others will require sustained political will and institutional courage.
Political leaders are an important driver in ensuring that military institutions learn from other people’s wars. Given the tight coupling between political and military institutions in peace and war, politicians must understand where change is occurring and then drive and resource change in political decision-making as well as in military institutions and other national security organisations. In the Australian context, this should include a mandatory annual report from the ADF to the government on changes in the character of war and the specific capability and doctrinal priorities that must be addressed as a result. The interventions that might be explored in such a report — including learning and adaptation in technology, force structure, doctrine, and personnel allocations — all demand political direction.
At the same time, politicians must be more exposed to the rapid evolution of military and national security technologies and strategies. This might take the form of civil-military wargames and enhanced intelligence briefings. The recent interim report of the Royal Commission on Antisemitism and Social Cohesion recommended that counter-terrorism responses might be improved in Australia by having the “National Security Committee ministers, including the Prime Minister, […] participate in a counter-terrorism exercise, along with all National Cabinet members, within nine months of each federal election”. A similar requirement for participation in war-management exercises and sustained strategic and technological briefings might be considered for political leaders. The war-management exercises might include exposure not only to military-technical change but also to the strategic lessons about will, agency, and the limits of coercion that modern wars have demonstrated.
The most important single change Australia’s defence establishment can make is not to its equipment holdings or force structure. It is to evolve its leadership philosophy and its promotion processes. The selection of the right leaders plays a central role in rapidly closing the gap between the emergence of new technology and its adoption by military institutions.
The ADF’s leadership selection and promotion processes, like those of most Western militaries, have been shaped by decades of stability operations and peacetime garrison management. Those processes tend to reward what one framework calls the “regulator” class — leaders who are disciplined, risk-averse, and skilled at navigating bureaucratic systems — over the “ratcatcher” class — leaders who are adaptive, intellectually restless, and willing to challenge established assumptions. The ADF should, as a matter of urgency, review its promotion pathways to identify whether they structurally disadvantage officers and NCOs who have demonstrated innovative and adaptive qualities.
The Chief of the Defence Force and each Service Chief should make explicit, public, and actionable statements about institutional tolerance for intelligent risk-taking and experimental failure. They should define the specific range of acceptable failures, name the learning and innovation activities they will resource and protect, and alter the promotion criteria by which future leaders are assessed.
This must be accompanied by reforms in Professional Military Education, which military scholar Michael Evans has characterised as suffering from “a lack of institutional understanding of the relationship between the health of the Australian profession of arms and the strength of the education system that underpins military activity. This lack of understanding incurs risk, since it is occurring at a time when the very concept of professionalism is being challenged by new forms of occupational expertise unleashed by the digital revolution.”
Finally, military leaders at all levels must demonstrate not only that they have learned, but also that they have created conditions in which their subordinates can learn.
The integration of AI into military learning and decision-making is not a future challenge. It is a present operational reality in Ukraine and the Middle East, and it is being developed at scale by many nations. The ADF should implement tailored analytical AI for strategic and operational functions to support learning and adaptation.
AI can help fuse disconnected learning processes, accelerate analysis, and speed up and enhance the quality of military adaptation and strategic decision-making. It can help identify relevant lessons in the vast volume of open-source, allied and classified reporting from Ukraine and Iran, and assist in translating those lessons into specific capability and doctrinal recommendations. The ADF should develop the institutional capacity to use AI as a learning accelerant, not as a substitute for human judgement but as a tool for improving the speed and quality with which that judgement is informed.
Ukraine’s Operation Spiderweb, in which civilian trucks were used to smuggle small drones deep into Russia and attack its airbases, demonstrated that modern nations are not only vulnerable to long-range missile strikes. Short-range drones can also be employed in areas that in previous eras may have been considered safe. The ADF’s limited counter-drone capability is a critical vulnerability that must be addressed as a matter of urgency. The Australian Government should commit to rapidly establishing dedicated drone and counter-drone units in each service and in the reserve components. These should be equipped with massed, cheap systems and not solely the small numbers of expensive platforms that currently dominate ADF acquisition planning.
The Ukrainian experience has demonstrated that the most important drone capability decisions are not about acquiring the highest-end system available. Key decisions are about quantity, speed of production, rapidity of software iteration, and the organisational capacity to integrate drone operations into joint and combined-arms planning at every level.
The Australian Government should invest in indigenous drone and counter-drone production capacity. Indigenous solutions support local industry, guarantee supply in wartime, and can often be delivered more quickly and cheaply than imported systems. Tax incentives for investment in Australian drone manufacturing should be legislated, alongside streamlined regulatory pathways for rapid fielding of new uncrewed systems. Australia already has companies producing counter-drone systems, rockets, rocket fuel, drones, and situational awareness systems. These should be nurtured as strategic industrial assets.
The analysis in Section 3 establishes the case plainly: Ukraine fielded new strike systems in months; equivalent Western programs take decades and cost an order of magnitude more. The question for the ADF is not whether to acknowledge this gap but how quickly it can close it. The answer depends on fundamental reform of how Defence acquires capability.
The Australian Department of Defence, with explicit direction and continuing oversight from Government, should develop and implement emergency acquisition pathways for urgent operational requirements. These pathways must accept higher technical risk in exchange for shorter fielding timelines. These should be modelled explicitly on Ukraine’s approach to rapid defence-industrial development: establishing clear operational priorities, empowering industry to experiment and fail quickly, and maintaining tight feedback loops between user and developer.
The ADF should study the Ukrainian Brave1 defence technology ecosystem to enable this effort. The Brave1 approach, where a government organisation plays the role of military-industry intermediary, has connected hundreds of Ukrainian defence companies, government procurement officials, and frontline users in a structured system for rapid fielding of battlefield innovations. The ADF should have an analogue suited to the Indo-Pacific context. This would require a significant reduction in the procurement bureaucracy’s risk aversion and a corresponding increase in the accountability of program managers for speed and operational relevance rather than process compliance.
The war in Ukraine since February 2022, and the more recent wars in the Middle East, have provided unprecedented learning opportunities. These wars are highly visible, documented, and widely analysed. The recent war in Iran has confirmed and amplified many of Ukraine’s lessons, particularly on the centrality of drone warfare, the inadequacy of Western counter-drone capabilities, the effectiveness of cheaper long-range strike systems, and the importance of cognitive warfare to modern war and strategic competition. And yet the response of Western institutions, including Australia’s, has been characterised by rigidity, inertia, and what can be called a humility deficit: an unwillingness to genuinely confront the implications of what is being demonstrated in real time on real battlefields.
For Australia, the consequences of continuing the current pattern will be measured not in budget line items but in the future performance of the ADF. The scenarios relevant to Australia — to protect Australian sovereignty and prosperity in a contested Indo-Pacific and on Australian territory against adversaries deploying massed autonomous systems — will all be environments in which the lessons of modern wars are directly applicable.
The recommendations in this paper are not radical. They do not require the abandonment of every existing capability program or wholesale restructuring of the ADF. They require something more difficult: a shift in institutional culture, a different kind of leadership, and the genuine humility to recognise that a military institution that learns more slowly than the threat evolves is already falling behind.
Reforming military organisations is hard under the best of circumstances. In the current era, where transformative technologies are changing the character of war as fast as in any time in history, the leadership and cultural challenge of organisational transformation is massive. But the ADF and the Australian Government cannot shy away from a task that will demand sustained focus over many years. It will require leaders who set the conditions for, and actively nurture, innovation. And as Martin Dempsey, former chairman of the US Joint Chiefs of Staff, has written of adaptation, “most innovation starts with a bit of rebelliousness; it’s worth considering how we will allow this into our own organisations”. This is good advice for the contemporary leaders of the Defence enterprise in Australia.
Note: All currency is quoted in US dollars.
