Defence industry is deterrence – and alliance supply is conditional

When NATO leaders met in Ankara in July 2026, industry was the centrepiece of the summit. The Strategy for Industry–NATO Cooperation, endorsed by all 32 allies, treats industrial capacity as an instrument of deterrence in its own right rather than merely a procurement support function. Australia was represented at the summit’s defence-industry discussions and at a meeting that included NATO’s four Indo-Pacific partners, where participants agreed to deepen cooperation in defence industry, cybersecurity and advanced technologies. But the summit’s industrial outcomes have so far drawn little analysis at home.

Four outcomes matter most. First, the NATO summit’s central innovation is to treat industrial capacity as an operational capability instead of an economic or procurement issue – that is, to treat it as something to be measured, exercised and improved rather than merely assumed. Secondly, it defines industry broadly: software, data, digital infrastructure and skilled people sit alongside production lines, and innovative small firms beside traditional major defence contractors. Thirdly, it commits to action, not description. NATO has publicly signalled demand through a consolidated, unclassified capability pipeline. It has tied industrial capacity to its capability targets and pledged to stress-test industrial surge through exercises. Fourthly, it puts collective weight behind all of this, including more than US$50 billion (A$71 billion) in announced procurements and a commitment to keep dismantling defence trade barriers among allies.

What does this mean for Australia?

First, if industry is an integral element of military capability, readiness and deterrence, it should be held to readiness standards the way armed forces are. The question is no longer whether an industry exists, but whether it can generate operational effect within the warning time and conflict duration defence planners assume. Measures of sufficiency, surge capacity, resilience and recovery are ultimately more useful than traditional assessments of industrial health, and NATO will now exercise them. Australia should do the same.

Secondly, the industry we should be preparing is broader than the one Australian policy usually pictures. It encompasses software, data, digital infrastructure, research institutions, skilled workforces, supply chains and small and medium enterprises, as well as traditional manufacturing. Its strategic value increasingly lies in adaptability as much as capacity. Success depends not only on how much industry can produce but also on how rapidly it can redesign, reconfigure, substitute components and incorporate new technologies. A policy that counts only factories will miss both dimensions.

Thirdly, the force needs both exquisite capability and producible mass. Conflicts in recent years have shown sustained demand for high-end interceptors and long-range precision weapons alongside cheaper munitions and drones employed in very large numbers. And building capability is only one of many life-cycle stages: maintaining, repairing, replenishing and upgrading are further stages requiring different industrial capacities, and they deserve separate assessment. Whatever Australia buys or builds, the non-negotiable is sustainment – a nation unable to sustain and adapt the systems it operates has effectively outsourced part of its freedom of action. Producible mass is where Australian industry can realistically grow. But the same scarce workforce, propellant and explosives production, and facilities will also need to support sustainment and, in some cases, production of exquisite systems, so sovereign investment choices are unavoidable.

Fourthly, defence industry is best managed as one interconnected system – the full network of firms, infrastructure, skills and supply chains, and the dependencies among them – rather than as individual sectors or companies. It also needs to be viewed from national and alliance vantage points at once. Turkey’s experience shows why the alliance vantage point matters. Its decades of embargoes and program exclusions, including removal from the Lockheed Martin F-35 program, have shown that supply from partners can become conditional when politics diverge. Turkey’s response was to build domestic capability in drones, missiles and combat aircraft, even though critical dependencies remain. For Australia – reliant overwhelmingly on the United States, with Britain’s industrial role growing through AUKUS – the same conditionality is a risk to be managed.

Fifthly, decisions around independent capability are choice of degrees, not binaries. For NATO members, those choices are complicated: national industrial policies, transatlantic supply relationships and an expanding set of European Union instruments with their own preference rules all pull at once. Australia’s situation is structurally simpler, but it is still a choice that should be made deliberately. Countries also face more than one such choice per system. For every aspect of capability, the practical question is which life-cycle stages (design, production, sustainment and upgrade) require domestic control and which can safely rest with trusted partners or the global market. Production can often be shared, but the capacity to keep systems operating and adapting often cannot. Framed this way, sovereignty stops being an all-or-nothing argument and becomes a series of deliberate, reviewable decisions for every system.

The NATO summit’s enduring lesson is that governments should begin by asking not what they should manufacture, but how much control they require over each element on which military capability depends. Turkey learned that lesson after losing access to allied capability. Australia can learn before it faces the same risk.