As warfare shifts decisively toward autonomous and distributed systems, Australia’s massive investment in nuclear submarines risks locking in a costly and inflexible strategy.

Richard Marles gave a polished performance at the National Press Club. Smooth, confident, practised. He spoke about drones, autonomy and the changing character of war. He sounded modern. But beneath the language of innovation sat an older and much heavier reality: Australia is still tying itself to one of the slowest, most expensive and least adaptable military projects in its history.

That is the contradiction at the heart of AUKUS.

Marles is not unaware of autonomous systems. He referred to uncrewed capabilities and undersea warfare. But that only sharpens the question. If Defence now openly accepts that autonomy, robotics and distributed systems are transforming warfare across all domains, why is Australia still betting so much on a tiny fleet of immensely costly, crewed nuclear submarines that will arrive deep into the 2030s and 2040s?

This is not a minor procurement issue. It is a question of strategic judgement.

The case for AUKUS rests on the assumption that large, crewed nuclear submarines will remain among the most valuable assets in the maritime battlespace for decades to come. That may prove true. But it may also prove disastrously wrong. The point is not that submarines are already obsolete. They are not. The point is that Australia is making an extraordinary long-term bet in a technological environment changing far faster than the program itself can adapt.

That should make any serious government cautious. Instead, Canberra behaves as if the matter is settled.

History offers a warning. In the early twentieth century the great powers poured fortunes into dreadnought battleships, symbols of industrial might and naval prestige. They were not instantly useless. But new technologies and new operational realities quickly eroded their dominance. Military establishments had invested too much money, too much institutional pride and too much doctrine to rethink the model in time.

That is the real relevance of the dreadnought analogy. Not that history repeats mechanically, but that states often mistake the apex of one era’s military technology for the foundation of the next.

That danger is obvious with AUKUS. Nuclear submarines are a mature technology. Autonomous undersea systems are not. They are improving quickly in endurance, sensing, networking and mission range. Underwater communications remain difficult. Swarming underwater is far harder than in the air. Human-crewed submarines will not simply vanish. But the direction of travel is unmistakable: the future undersea battlespace will be more distributed, more robotic, more sensor-saturated and more hostile to concentrated, exquisite platforms.

In that context, the problem with AUKUS is not merely cost, although the cost is staggering. It is rigidity. Australia is locking itself into a force structure designed around a handful of elite assets whose strategic rationale may narrow before they even enter service. The more expensive the platform, the more reluctant commanders will be to risk it. The more politically symbolic it becomes, the less usable it may be in practice.

A submarine that cannot be risked is not a deterrent. It is a shrine.

This is where the drone discussion matters. Marles is happy to speak about drones as an addition to the force, as a supplement, as evidence that Defence is keeping up with change. But what if they are not a supplement? What if they are part of a broader shift that should force a rethink of the whole investment model? What if the lesson of autonomous warfare is not “buy some drones as well” but “stop spending so much on vulnerable prestige systems with decades-long delivery horizons”?

That is the debate Canberra does not want.

Because once that question is asked seriously, AUKUS stops looking like strategic vision and starts looking like strategic inertia dressed up as alliance management. It begins to resemble what it has always partly been: not simply a defence project, but a political project, a way of binding Australia more tightly to Washington, advertising resolve, and cloaking dependence in the language of technological sophistication.

The tragedy is that Australia does need a serious debate about maritime defence. We do need to think hard about sea denial, choke points, infrastructure protection, autonomous systems, missile strike, industrial capacity and strategic geography. We do need to prepare for a harsher region. But none of that requires blind faith in a single gold-plated answer.

Indeed, the lesson of technological disruption is the opposite. In periods of rapid change, resilience comes from adaptability, diversity and strategic humility. It does not come from placing one colossal bet and then building the national security establishment around defending it from criticism.

That is why Marles’ performance was so revealing. He spoke the language of future warfare while defending the procurement logic of the past. He praised disruption while protecting orthodoxy. He acknowledged change without following its implications to their conclusion.

Australia may yet acquire nuclear submarines that serve a useful role. But that is not enough to justify the present scale of commitment. For nearly $368 billion, Australians are entitled to more than polished rhetoric and alliance piety. They are entitled to ask whether this fleet will still make sense in the world for which it is actually being built.

That question is not anti-defence. It is the essence of defence thinking.

And the more insistently Canberra avoids it, the more AUKUS looks less like insurance against the future than a monument to an ageing strategic imagination.