From Canberra to Tokyo: Undersea Power and the Nuclear Submarine

10/27/2026
By Robbin Laird

Within a single week in October 2026, two allied capitals made overlapping arguments about how a maritime democracy should build military power in the Western Pacific.

In Australia, Professor John Blaxland of the Australian National University made the case for AUKUS twice in two days. He did so first in an article with Kate Rosengreen in ASPI’s The Strategist (The Strategist), and then in a 69-minute interview on the Devil’s Advocate podcast. In Tokyo on 23 October, Defence Minister Shinjiro Koizumi told the Diet that the question of nuclear-powered submarines is open. Later that day he announced that the research laboratories of the Acquisition, Technology and Logistics Agency (ATLA) will be reorganised around warfighting functions rather than platform types (Kyodo via Daily Sports; Sankei).

The two discussions were separate. Read together, they show that Blaxland’s case for undersea power is not only an Australian argument about an Australian purchase. It is part of a wider regional rethinking of how allied capability should be organised, and Japan’s remarks both support that case and complicate it.

Strategy, Not a Purchase

Blaxland and Rosengreen open with a reframing: “AUKUS is not just a submarine acquisition program; it is a long-term strategy for strengthening Australia’s agency in a contested Indo-Pacific.” They define undersea power as “the national capacity to understand, monitor, protect and contest the undersea environment,” including cables, seabed resources and maritime energy infrastructure.

The definition changes what counts as success. If AUKUS is just a boat buy, every delay becomes the whole story. If it is the building of a national capability, the question is whether Australia is getting better at seeing and contesting the undersea domain, and building the people and industry to keep doing so.

Koizumi’s laboratory announcement makes the same move inside a bureaucracy. ATLA currently has separate institutes for air, ground, naval and new-generation equipment. Koizumi said these would be rebuilt around AI, cyber, command and control, and information and communications (Sankei). He described AI and cyber as “synonymous with new ways of fighting” (Mainichi via Yahoo! News). He also set out a goal that has nothing to do with any single platform: “We must achieve the ability to find, decide and respond faster than the adversary.” Nikkei’s report, as relayed in a summary posted on X, said the aim is to apply research across platforms and get results into the force faster, with the reorganisation targeted within fiscal year 2027 (summary of Nikkei report on X).

Both men reject the same assumption: that capability is best understood and organised one platform at a time. Blaxland rejects it in how he frames a national strategy. Koizumi rejects it in how he structures defence research.

The Kill Web Made Institutional

The overlap is clearest on the kill web. Asked on Devil’s Advocate whether uncrewed systems offered a cheaper alternative to nuclear-powered submarines, Blaxland rejected the premise: “not either/or… it’s everything all at once.” He described the shift from the kill chain to the kill web. In a kill web, crewed and uncrewed platforms across air, surface, subsurface, land, cyber and space are networked so that missions can be handed from one to another.

That is a doctrinal claim. Koizumi’s reorganisation is an attempt to build the institution such a doctrine requires. Laboratories organised around aircraft, ships and ground vehicles tend to produce better aircraft, ships and ground vehicles. Laboratories organised around AI, cyber, command and control, and communications are designed to produce the connective tissue that lets a distributed force act as one. “Find, decide and respond faster” is a sensing-and-decision formula, not a platform formula.

In a kill web, a nuclear-powered submarine’s value lies less in what it can do alone than in how far it extends a network of sensors and effectors, and how long it can stay there. That only works if the network exists, and building it is as much a research and organisational task as a shipbuilding one. Japan is now restructuring its defence research on that basis.

Why Nuclear Propulsion: Two Arguments

On submarines, the overlap is real, but the reasoning differs in an instructive way.

Blaxland’s core technical case is about detectability at the surface. He does not claim the oceans are becoming transparent. Below the surface, they “remain opaque.” The vulnerability is the snort. A conventional submarine must raise its breathing pipe to run its diesels and recharge its batteries, and a transit from Perth to Darwin requires that “a dozen or more times.” Pattern analysis, AI-driven anomaly detection, proliferating drones and low-earth-orbit satellites make that time at the surface “detectable to a degree that threatens survival in conflict.” Blaxland adds a geographic point. A conventional boat may suit a navy operating close to home, but it is badly matched to one whose operating areas begin thousands of kilometres from its bases.

Koizumi did not make that argument. At the House of Representatives Security Committee, he took a positive view of discussing nuclear propulsion as part of the revision of Japan’s three security documents, due by year-end. He said it was “important to discuss all other forms of propulsion, including comparisons” (Kyodo via Daily Sports). His reasons were speed, submerged endurance and the conditions crews live in. He noted that power limits on the Maritime Self-Defense Force’s diesel submarines mean sailors cannot shower every day.

The difference is worth noticing. Japan operates much closer to its own approaches than Australia does. By Blaxland’s geographic logic, Japan is the kind of navy for which conventional boats should be well suited. Yet Japan’s defence minister now wants nuclear propulsion considered as part of the formal strategy revision, and he grounds the case in operational endurance and crew conditions rather than prestige.

That suggests the pressures Blaxland describes are not purely Australian. Submerged endurance matters for persistent presence along Japan’s long southwestern island chain as well as across the Indian Ocean approaches. Koizumi did not mention detectability, and his remarks should not be read as endorsing Blaxland’s surface-vulnerability argument. But two allies with very different geographies are now asking whether the conventional submarine still meets the demands of persistent presence in a contested Western Pacific.

Koizumi was also careful about what he did not say. He stressed that the government must “deepen the examination, including merits and demerits, without deciding the matter in advance.” Asked about nuclear deterrence policy, he repeated that the government “does not rule out any options” (Kyodo via Daily Sports). This is an opening, not a decision, and it is a long way from Australia’s committed program.

People as Combat Capability

The detail about showers may seem trivial. It is not. It points to the constraint that runs through both countries’ debates: people.

In the Devil’s Advocate interview, Blaxland acknowledged that Australia needs engineers faster than domestic training can supply them. On balancing the need for talent against the demands of security vetting, he called it “a million dollar question” and offered no clear answer. Koizumi’s remark comes from the other end of the same problem. In a force already short of people, the conditions under which sailors live and work affect recruitment, retention and how long a crew can sustain operations.

In both cases, people rather than hardware are the binding constraint. A propulsion choice is also a workforce choice, and neither Canberra nor Tokyo has yet shown how it will close the gap.

Denial, Deterrence and Allied Signals

Koizumi said “denial capability” will be a keyword for the revised security documents: Japan “will not attack, but must establish the power to deny threats approaching it” (Mainichi via Yahoo! News). That closely matches Blaxland’s view of deterrence. He presented AUKUS as insurance for deterrence, not preparation for war. Wars start when an aggressor thinks victory will be easy, so Australia should be a “porcupine” or a “prickly pear” that makes a potential adversary “think twice.”

Blaxland also argued that declaring Australian neutrality on Taiwan in advance would have a “negative demonstration effect” on Japan and the Philippines: allied signals reinforce or undercut one another. Japan’s October statements send a signal in the other direction. Whatever Washington’s political messaging, which Blaxland called “deeply alarming,” the regional allies are still investing in denial.

Where the Models Diverge

The comparison also shows a real difference. Blaxland describes AUKUS as “a transition from interoperability to interdependence.” Its goal, in his and Rosengreen’s words, is “resilient sovereignty within an interdependent alliance network, not autarky.” Submarine Rotational Force–West at HMAS Stirling is the clearest example: Australia gains capability and the US Navy gains an Indian Ocean base.

Koizumi’s industrial remarks point another way. He said the government plans legislation in next year’s Diet session to support the defence industry more strongly than before. Sankei read this as pointing towards stronger domestic production and an independent agency to promote research and development (Sankei). Japan already designs and builds its own submarines. It begins from a much stronger sovereign industrial base than Australia did, and its instinct is to strengthen it further.

If Japan’s examination of nuclear propulsion becomes serious, it will test whether the AUKUS model is a one-off or a template. Australia chose interdependence partly because, as Blaxland notes, Australians want independence but will not pay for it. Japan may weigh that choice very differently.

The Common Test

Blaxland and Rosengreen set the benchmark for AUKUS. By the mid-2030s, the question is whether Australia has “the national systems, people, infrastructure and industrial capacity required to make and implement its own decisions to exercise undersea power.” That is a higher bar than delivering submarines, and the right one.

Japan faces a parallel test. Rebuilding laboratories around functions is a declaration of intent. Whether it changes anything depends on whether budgets, career paths and acquisition programmes follow the new structure, or whether the platform communities survive under new names.

The common lesson is clear. In both capitals, the debate is moving away from the platform and towards the system the platform serves: the network that finds, decides and responds, and the people and industry that sustain it. Blaxland has made that case for Australia. Koizumi’s remarks suggest Japan is starting to make it for itself.

Sources