America Promised Australia Submarines It Can't Yet Build: The AUKUS Industrial Gap

Executive Summary

The United States operates around 49 nuclear-powered attack submarines. At any given moment roughly 18 of them cannot leave port — tied up waiting for maintenance the shipyards cannot get to. Hold that number, because under AUKUS Pillar 1 the US has committed to transferring three to five of its newest Virginia-class boats to Australia from the early 2030s. On current production, it cannot build them without drawing down a fleet that is already roughly a third sidelined.

That gap — between the promise and the steel — is the single most important number in the Pacific right now. The usual coverage argues about politics: will Australia pay, will Washington stay committed across administrations. Based on more than a decade as a British Army intelligence analyst, I assess those are the wrong variables to anchor on. The variable that decides AUKUS is industrial: how many submarines the United States can physically build and keep at sea per year. Throughput, not intent.

---

The lens: a single pipe of fixed diameter

Only two yards in America build these boats: General Dynamics Electric Boat and HII's Newport News. Only a handful of public yards maintain them. Picture a single pipe of fixed diameter through which everything — new construction and repairs — has to flow. Money can fill the pipe. Political will can demand more flow. Neither widens the pipe inside a few years, because that takes skilled welders, a casting and component supply chain, and most of all, time.

Two honest caveats before the case. These figures predate the FY2027 appropriations now clearing Congress, and a genuine sustained surge could bend the trajectory. And the build rates here are the Navy's own stated targets, not independently audited delivery. So here is what would change the assessment: sustained delivery past 1.5 boats a year by 2027, or a funded second submarine lot breaking ground. Absent those two, the structural read holds.

Before America builds a new boat, it can't use a third of the ones it has

In FY2024, 16 of the Navy's 47 attack submarines — 34% of the fleet — were in depot maintenance. The Government Accountability Office found the Navy spent roughly $4.2 billion over the last decade just crewing and supporting submarines that couldn't deploy. The cause isn't combat losses. It's that America has only four public shipyards to fix these boats, and they are old, overloaded and short-staffed.

The case that crystallises it is the USS Boise — an attack submarine that sat awaiting repair so long it managed barely a single patrol in over a decade. After roughly $1.6 billion spent trying to return her to service, the Navy gave up and inactivated her. A front-line boat written off not by an adversary, but by a maintenance system that couldn't keep up. The tempting counter is "more money clears the backlog." But the limiter is yard capacity, not the cheque. You cannot surge overhaul throughput inside this window any faster than you can surge new construction — and for the same reason.

The build numbers don't reconcile

The United States today delivers about 1.3 Virginia-class submarines a year. The Navy target is two. In 2023 the Navy said it would hit two a year by 2028; this month the Chief of Naval Operations told appropriators it would be 2032 — a four-year slip to the right.

Be precise about why that matters, because "slip" sounds like one bad stretch. It isn't. The US has been planning and funding two Virginias a year since 2011, and in all that time it has never actually delivered two in a single year. Since 2022 the real rate has run around 1.1 to 1.2. This is not a programme that stumbled once; it is a programme whose required rate and whose delivered rate have never met. That is the structural fact, and it is why a near-term recovery rates low.

Now layer AUKUS on top. Pillar 1 commits the US to transfer three to five Virginias to Australia from the early 2030s. The figure that should focus the mind is the Navy's own: to sell those boats to Australia without drawing down its own fleet, the industrial base needs to build about 2.33 attack boats a year — plus one Columbia-class ballistic-missile boat. Hold the numbers together: 1.3 today; a two-a-year target missed for 13 years running; 2.33-plus-a-Columbia to honour AUKUS. That gap is close to double what the yards have ever delivered.

The will and the cash are present — so what's binding?

The easy story is that AUKUS is at risk because someone won't pay or won't commit. The data says the opposite. Australia is funding its half — a roughly $40 billion (USD) defence budget with an AUKUS ramp — and hedging like a serious customer, with a near-$8 billion life extension on its existing Collins-class boats and around $3 billion going directly into the American industrial base. The will is present, the cash is present, and the hedge is funded.

What's binding is priority and people.

Priority. The two yards everything depends on have a boat that outranks Virginia: the Columbia-class ballistic-missile submarine, the sea-based leg of America's nuclear deterrent and the Navy's number-one shipbuilding priority, full stop. The first is tracking to 2028 with no schedule margin. Columbia and Virginia draw on the same two yards, the same welders, the same suppliers — so when the priority boat needs capacity, it gets it, and Virginia is the variable that gives. The deterrent crowds out the attack fleet by design.

People. The latest plan adds billions to the submarine industrial base — roughly $9.8 billion flowing in, including a $6.2 billion addition, plus the Australian contribution — but that money buys capacity on a multi-year delay. The Secretary of the Navy says the sector needs to hire something like 250,000 workers this decade. Electric Boat alone planned to add 8,000 in 2026, but the year before it had to cut its 2025 hiring target to about 3,000, because its suppliers couldn't keep pace with the boats already on order. That's the bottleneck behind the bottleneck: you can hire welders faster than your vendor base can deliver castings and components, and then your new hires have nothing to fit. A skilled submarine welder takes years to qualify. A sole-source casting supplier takes years to stand up. Neither responds to a supplemental inside the AUKUS window.

Allies help at the edges but cannot widen the pipe. The UK's Babcock is building US submarine components at Rosyth; Australia is putting billions into the American base. That eases the supply chain. But final assembly of a nuclear-powered submarine happens at two American yards under US law, and no amount of allied funding adds a third slipway or a fifth public drydock inside a decade. Allied help buys resilience at the margins, not throughput.

The fine print: the transfer is conditional

From the early 2030s the US transfers three to five Virginias to Australia — the first nuclear-powered submarines that country has ever operated — while Australia and Britain build the SSN-AUKUS class for the longer term. But the transfer is conditional: the United States retains the right to withhold the boats if handing them over would erode its own undersea capability at the time.

Set that clause against everything above — a fleet already a third sidelined, a build rate that has never hit its target, and a deterrent boat with first call on the yards — and the conditionality clause stops looking like boilerplate. It is the exact lever the industrial gap makes most likely to be pulled. The opportunity cost is concrete: at about 1.3 boats a year, three to five Virginias is roughly two and a half years of total US attack-submarine output. Those are the same boats that would otherwise be the margin keeping the US fleet above its own floor. The shortfall the US opens with, and the boats it is promising Australia, are in large part the same boats counted twice.

Three branches, each with an indicator to watch

Branch 1 — the surge lands, delivery holds on schedule. Unlikely (15-30%). Assumes the appropriations fund a sustained two a year and the yards finally resolve the supplier and workforce bottlenecks they never have before. Watch: deliveries pushing past 1.5 boats a year by 2027, plus a funded second-yard expansion actually underway.

Branch 2 — a partial slip. The likely path (45-60%). The boats come, but late into the mid-2030s, and the US fleet dips below its own floor for a period. Assumes a modest throughput gain and enough political pressure to honour AUKUS with the timeline stretching to absorb the shortfall. Watch: Virginia deliveries plateauing around 1.3-1.5, and Australia extending the Collins life-of-type programme again as a bridge.

Branch 3 — the transfer fails or defers. A realistic possibility (20-35%): low probability, high consequence. Australia bridges with Collins and a sovereign build. Assumes Washington prioritises its own fleet under Chinese pressure and invokes the delivery-conditionality clause. Watch: US signalling the opt-out, talk of renegotiating AUKUS, or Canberra accelerating a sovereign-build conversation.

The most likely outcome is neither collapse nor success. It's late.

Why this runs beyond three submarines

US naval intelligence assesses Chinese shipbuilding capacity at something like 200 times that of the United States. Be precise: that's a tonnage-capacity figure from a 2023 slide the Navy itself calls non-definitive, not a 200-to-1 count of warships. But it tells you the domain where the gap is widest is exactly the domain AUKUS is trying to compete in. So the third judgement, held at moderate confidence because it concerns perception: if the US visibly slips on Australia's boats, that slip gets read in Beijing and in non-AUKUS allied capitals — Tokyo, Seoul, Manila — as evidence that American guarantees outrun American capacity. Once partners price that in, the effect runs well beyond three submarines. That is plausibly the dominant Indo-Pacific signal of the next 24 months.

The verdict, and what to do with it

The transferable lesson outlasts the programme: industrial capacity is a promise you make decades in advance. You cannot surge it, you cannot buy it on a deadline, and you cannot transfer your way out of a shortfall you never fixed. Allied commitment can be signed in an afternoon; the steel to honour it takes a generation. That mismatch — fast politics, slow industry — is the structural fault line under the Pacific, not just under AUKUS.

The single indicator that resolves all of it is the Virginia delivery rate in 2027. Past 1.5 boats a year, the promise is recoverable. Still stuck near today's 1.3, that's the moment to re-cross every assumption that depends on AUKUS arriving on time.

---

This piece is the written companion to the Iron Command deep-dive on the AUKUS submarine gap. For the full assessment with every figure graded, watch the video above or read it each week in the Pacific Brief. If you allocate capital or underwrite risk against the Indo-Pacific, that timeline-pricing work is what Iron Command Advisory does.