Fight Tonight, Prepare for Tomorrow: The ADF and the Age of Autonomous Warfare
“The comfortable certainties of the post-Cold War era are dissolving before our eyes.”
That observation, made by ACM (Retired) Mark Binskin at the Sir Richard Williams Foundation’s September 2025 seminar, resonates now with even greater force as the Foundation prepares to convene again on April 23, 2026 at the National Gallery of Australia.
The theme this time — Fight Tonight: Exploiting Australia’s Strategic Advantage — takes the same foundational question and sharpens it to a fine edge: not simply whether Australia can fight, but how it can fight tonight, with the force it has, exploiting the geography and industrial base it possesses, against a pacing threat that is accelerating.
I return to Australia for this seminar carrying two books that bracket the very question at the seminar’s heart. The first, Fight Tonight Force: Combat Readiness at the Speed of Relevance, emerged from the Foundation’s September 2025 seminar.
The second, Lessons from the Drone Wars: Maritime Autonomous Systems and Maritime Operations, examines what the wars in Ukraine and the Middle East have actually demonstrated about uncrewed systems, autonomous capability, and the gap between technological hype and battlefield reality.
Taken together, they frame a key strategic argument: Mastering the fight tonight is not a distraction from building tomorrow’s autonomous force. It is the essential precondition for it.
The Fight Tonight Imperative
The September 2025 seminar established a foundational tension that no subsequent discussion of Australian defence can ignore. On one side sits the imperative of readiness, the force in being, deployed tonight, capable of generating deterrent effect before a single additional dollar of procurement has been spent. On the other sits the demand for transformation, the need to build mass, depth, agility, and industrial resilience for a conflict that may not come tonight but is shaping now.
ACM Binskin’s framing from September 2025 captures the dilemma precisely: “Modern conflicts are won as much in the information space as the traditional battlefields,” and democracies that wait for perfect preparation risk ceding the initiative entirely. Professor Justin Bronk of RUSI, who returns to the April 2026 program as its marquee speaker on global trends in air and missile defence, placed the timeline question in starker terms at the September event: the modern threat spectrum demands a fundamental rethink of air defence, and “the timeline for preparation is far shorter than most assume.”
Against this backdrop, the April 2026 seminar’s four strategic themes, building combat mass and depth, generating tempo across domains, enhancing industry capability, and surviving to operate, represent a coherent progression. They begin with the force in being and work outward toward the national support base, the industrial foundation, and ultimately the question of national will. Colonel David Beaumont’s observation from the earlier seminar that success “depends not merely on having the right equipment, but on maintaining the flexibility to rapidly reconfigure logistic networks, repurpose existing capabilities, and sustain operations across vast distances with uncertain supply lines” sets the terms for the industrial and logistics discussions that will dominate the afternoon session on April 23.
What the Drone Wars Actually Teach
Into this strategic architecture steps a contribution that deserves particular attention from the April 2026 audience: the argument made by Wing Commander Keirin Joyce in both the foreword to Lessons from the Drone Wars and in his April 15, 2026 piece in The Strategist, titled “ADF must master the fight tonight before betting on tomorrow’s autonomy.”
Joyce writes from a position of rare credibility, a serving ADF officer, former ASPI senior analyst, and deep researcher into autonomous systems. His argument is a bracing corrective to the hype cycle that has colonized much defence commentary on drones.
The core claim: the multi-rotor flight control systems and open-source software that define today’s FPV drone battlefield rest on a foundation of scientific research that peaked between 2011 and 2015. That is why Ukrainian drone operators could achieve scale militarization by 2022. The science was settled, robust, and open-source long before the first shot was fired.
By contrast, while research on autonomous, AI‑driven drone systems has expanded rapidly, much of the most celebrated work remains confined to controlled environments rather than combat. Recent advances such as the University of Zurich Robotics and Perception Group’s high‑speed autonomous racing drones demonstrated in competitions against elite human pilots since around 2020–2022 show that AI can beat humans on carefully designed race courses with known gates and constrained dynamics, but not yet in the complexity of a battlefield. Drawing on historical lags between peaks in enabling research and mature operational deployment,
Joyce argues that robust, reliable autonomy capable of independently engaging targets at scale is unlikely to be fielded in meaningful numbers before roughly 2026–2028 at the earliest, and potentially later. This is consistent with leading researchers’ emphasis that transferring such performance from simplified test ranges to cluttered, adversarial, and partially observable warzones is a long‑term challenge, not an imminent reality.
What passes for AI-autonomous strike drones in current field reporting, Joyce argues, are largely bootleg adaptations, companion computers hacking flight controllers to mimic pilot thumb inputs, keeping a target centred in the frame. Innovative, but brittle. Prone to oscillation. Easily confused by complex environments. Often less accurate than a skilled human pilot. The field claims of 80 to 100 percent hit rates should be viewed with deep scepticism. The ADF, Joyce argues, risks falling into what Gartner calls the “trough of disillusionment”, spending heavily on immature autonomous systems, experiencing operational failure in complex contested environments, and then cutting funding precisely when the technology is beginning to mature.
This argument sits squarely within the analytical framework I developed in Lessons from the Drone Wars. As I wrote there, there is a critical distinction between true autonomy where the device is capable of making independent decisions and the more immediately achievable realm of collaborative systems that amplify human judgment rather than replacing it. The Ghost Bat program, as both Joyce and I have argued, represents exactly this intermediate opportunity: in its current ISR configuration it can build the experiential knowledge base that will eventually enable more sophisticated autonomy, while delivering immediate operational value. The ADF should not wait for Level Five autonomy before integrating what Level Two and Three systems can do today.
The Human Element as Strategic Asset
The September 2025 seminar generated a ten-point synthesis of “answers to the speed of relevance,” and running through nearly all of them was a common thread: the human dimension is the ultimate constraint. Not platforms, not software, not procurement cycles — but people. Their training, their judgment, their ability to adapt faster than any adversary can observe, orient, decide, and act.
This is exactly Joyce’s point about the fight-tonight drone reality. The FPV drone capability that has reshaped modern warfare is not primarily a hardware story. It is a human skill story, a globally networked community of operators sharing code, tactics, and hardware fixes in real time, drawing on competitive racing leagues and open-source simulators to compress training timelines that traditional military institutions would measure in years. The first Military International Drone Racing Tournament was held in 2018. By 2022, those skills were on active battlefields. The institutional implication is profound: the ADF needs to treat operators as the capability, not merely the operators of the capability. Formalizing and scaling pilot training through the same open-source simulators and competitive structures that produced commercial drone racers is not a nice-to-have. It is a strategic investment.
AIRCDRE Peter Robinson, Commander of RAAF Air Combat Group, addressed this tension at the September 2025 seminar, discussing how the RAAF works to be ready for the fight tonight while preparing for tomorrow. Chief of Air Force AM Stephen Chappell’s “four Ds” (degrade, disrupt, destroy, defeat) and “six Cs” (capability, credibility, comprehended, communicated, collectively, and consistently) framework, most prominently displayed with the MQ-28 Ghost Bat, captures the institutional logic: near-term readiness and long-term transformation are not competing priorities but complementary ones, maintained in what Chappell himself calls a “healthy tension.” The April 2026 seminar, with Chappell again delivering the Chief of Air Force perspective, will test whether that tension has been productively maintained or whether one side has begun to dominate at the expense of the other.
Geography as Strategic Depth
One of the more underappreciated arguments to emerge from the September 2025 discussions, and one that the April 2026 seminar explicitly foregrounds, is the idea of Australia’s natural strategic advantages as a form of capital. The seminar’s aim, to explore “ways and means of exploiting Australia’s enduring natural strategic advantages to provide deterrence against attacks on defence assets and critical national infrastructure”, points toward something that distinguishes Australia’s strategic situation from almost every other major U.S. ally: the continent itself is a strategic asset.
Professor Stephan Fruehling of the Australian National University made the point in my interview with him last year that Australian strategic thinking has been drifting toward an uncomfortable middle, too dependent on U.S. alliance frameworks to develop genuine autonomy, yet insufficiently institutionalized within those frameworks to benefit from their full depth. His concept of “deliberate incrementalism” as a management framework for this tension is worth carrying into the April discussions.
RADM Brett Sonter’s concept of “security clusters”, enhanced collaboration between crewed and uncrewed systems as a template for regional maritime security that scales outward to partner nations, represents one practical expression of how geography becomes operational advantage. The April 2026 seminar’s maritime perspective, delivered by RADM Christopher Smith, Commander Australian Fleet, will have the opportunity to take that concept further. The sea-air integration question that runs through both Fight Tonight Force and Lessons from the Drone Wars, how distributed autonomous systems change the calculus of maritime power projection across the vast distances of the Indo-Pacific, sits at the heart of what the April discussions need to resolve.
The Industrial Foundation Question
If the September 2025 seminar established the intellectual architecture, the April 2026 seminar is where that architecture meets the constraints of fiscal reality. The seminar documentation is explicit: “acknowledging the current resource pressures in the defence budget, the seminar series will reinforce the need to balance near-term decisions in air and space capability to ensure Defence and industry investment also provides the sustainable foundations for future force structure planning and growth.”
Matt Jones of BAE Systems put the industrial mobilization question in historical terms at the September event: “The choice is stark: begin serious industrialization now during relative peace or face the consequences of unpreparedness when strategic patience runs out.” AVM (Retired) Robert Denney reinforced the point: the fight-tonight force must be prepared to “fight tomorrow night, next week and next month,” which demands that government and industry go further in collaborating on guided weapons and explosive ordnance manufacturing in Australia.
Joyce’s argument from The Strategist piece connects directly to this industrial debate. The value-generation opportunity for Australian defence industry, he argues, lies not in replicating low-margin hardware that can be bought cheaply from commercial vendors, but in mastering high-fidelity simulation and creating the global training sets that will eventually allow graduation from bootleg adaptations to resilient full autonomy. The barrier to full autonomy is not hardware: it is data. This reframes the industrial investment question: rather than competing with Chinese commercial drone manufacturers on unit cost, Australian industry should be investing in the data infrastructure that will define the next generation of autonomous capability.
The Two-Pronged Strategy
The analytical convergence between Fight Tonight Force, Lessons from the Drone Wars, and Joyce’s Strategist piece points toward a coherent two-pronged strategy.
The first prong is immediate and human-centric. Invest in the operators. Scale pilot training through open-source simulators and competitive leagues. Treat the operator cadre as the strategic asset, not merely the user of the asset. Equip them with the best available analogue and digital links capable of fighting through electronic warfare jamming that exists now, today. Build the resilient supply chains, particularly for the commercial-off-the-shelf components that define current FPV capability, that can sustain operations under attrition. This is the fight-tonight force, and it requires investment that is immediate, concrete, and unambiguous.
The second prong is future-oriented and data-focused. While fighting with humans today, prepare the data infrastructure for tomorrow’s machines. Commission the high-fidelity simulation environments. Begin generating the diverse machine-learning training sets that will teach autonomous systems to navigate complex, unpredictable environments, flying through urban rubble, operating under forest canopy, maintaining guidance through battlefield smoke, surviving GPS denial. Build the doctrine frameworks that will govern human-machine teaming as autonomy levels progressively increase. This is not a procurement decision. It is an investment in intellectual capital that will determine whether Australian autonomous systems, when they arrive, are genuinely robust or merely a more expensive version of today’s bootleg adaptations.
The Seminar in Context
The April 23, 2026 seminar at the National Gallery of Australia brings together a speaker list that reflects exactly the breadth this two-pronged challenge demands. Mike Pezzullo on whole-of-government security challenges. Justin Bronk on global trends in air and missile defence and strike. The Commander Air Combat Group. A USINDOPACOM liaison officer. The Australian Signals Directorate; and ultimately the Chief of Air Force himself. The afternoon’s panel, facilitated by AIRCDRE Jo Brick and spanning RUSI, USINDOPACOM, Space Command, and ASD, will have the opportunity to stress-test the strategic assumptions that will shape Australian defence investment for the coming decade.
What the September 2025 seminar established and what Fight Tonight Force documented was that the foundational questions of readiness, mass, industrial mobilization, and human capability are not separable from the transformational questions of autonomy, AI, and uncrewed systems. They are the same question viewed from different time horizons. Joyce’s essential argument, which I would commend to every participant at the April seminar, is that the hype cycle threatens to invert that relationship, to treat the transformational future as the real priority and the present fight-tonight capability as a placeholder. That inversion is strategically dangerous.
The democratization of air power is real. Cheap, commercial off-the-shelf technology does now allow non-state actors and smaller countries to generate precision effects previously reserved for major powers.
But the removal of humans from the loop, the genuinely autonomous, AI-driven swarm operating independently in GPS-denied, EW-contested, visually complex environments, is not yet a reality. The ADF that bets its near-term capability investment on that future, while neglecting the operator cadre, the training infrastructure, the supply chains, and the doctrinal frameworks that define today’s fight, is an ADF that risks losing the war it actually faces in order to prepare for one that has not yet arrived.
Fight tonight. Prepare for tomorrow. The Sir Richard Williams Foundation has been asking this question with growing urgency across two years of seminars. The April 23 gathering is an opportunity to convert that urgency into a coherent national strategy.