![COLUMN | Back on track? US icebreaker construction; Anduril invests in naval submarine manufacturing; war and how it creates business opportunities for billionaires [Offshore Accounts]](http://media.assettype.com/bairdmaritime%2F2026-10-12%2Fhcfestdn%2FUntitled-May-13-2026-at-09.00.32-1.jpeg?w=480&auto=format%2Ccompress&fit=max)
Last week, we looked at some of the recent deals in offshore supply that have been powered by industry-wide optimism in oil and gas, as Brent crude continues to trade around US$105 per barrel and the war in the Middle East rages on. The deadly Houthi attack on Riyadh's King Khalid International Airport this weekend showed that the world’s former second largest oil producer is still under heavy pressure, and whilst the number of ships transiting the Strait of Hormuz has risen, it remains under pre-war levels. The transit remains dangerous, and expensive, as war insurers price in the risk of Iranian attacks on vessels trying to cross on the Omani side under the cover of darkness and under the protection of the US Navy.
This week, China announced it will suspend all fuel exports this month to keep its domestic market supplied, and whilst President Donald Trump has announced a waiver on American sanctions on the import of Russian diesel, it is not clear how much diesel Russia can actually export given the Ukrainian drone attacks on its refineries have led to domestic shortfalls, characteristically Russian fist-fights at station forecourts and a ban on exports from Russia through to the end of October. Mongolia, which depends on Russian imports for 97 per cent of its fuel requirements under an intergovernmental treaty, has already run dry, prompting demonstrations in Ulan Bator.
Consumers are buckling under the pressure of higher fuel costs, but one group is making out like bandits: defence contractors.
Whether they will deliver on the promises they are making to governments remains to be seen.
In 2025, when the excitement of the golden renaissance of the American shipbuilding dream was at its highest, spurred by optimism from Washington and maritime influencers, we were party poopers (surely not?).
We pointed out that the problems American shipbuilder Bollinger Shipyards was facing delivering the polar security cutters (PSCs), big icebreakers under construction at Pascagoula, Mississippi, for the US Coast Guard, showed that it would take more than press releases and mission statements to revitalise American shipbuilding.
A country that struggles to stitch luxury handbags might struggle to produce complex technically advanced vessels of a type no American yard had built in more than four decades, after Lockheed delivered USCGC Polar Star and sister USCGC Polar Sea in the late 1970s. USCGC Storis, formerly Edison Chouest’s Aiviq, which was delivered from Houma Louisiana in 2012 at a price estimated at US$200 million for a Shell drilling support contract, and which the coast guard acquired in 2024, is a 16,240kW medium icebreaker.
These new PSCs being built by Bollinger, like Polar Star, are deemed “heavy icebreakers” with planned 33,700kW Caterpillar powerplants enabling them to break ice up to 1.8 metres thick at a speed of three knots. The original contract was awarded in 2019 to Halter Marine, which Bollinger acquired in 2022, but the first vessel of the three budgeted and planned was… delayed. Badly.
When we wrote the piece last year, the first PSC was not expected to be delivered until 2030, it was forecast to cost around US$2 billion, twice the budgeted original cost, and the second and third ships in the class were not yet even under construction.
With the keen eye of Secretary of Defense Pete Hegseth and his crack team of naval advisors on the job of reviving the industry as preparation for the construction of the new series of proposed Trump-class battleships, surely things would be quickly brought back on track? What could possibly go wrong?
This June, we got our answer, as per Josh Luckenbaugh in National Defense:
“The Government Accountability Office’s annual assessment of Homeland Security Department programs released in June painted a grim picture of the coast guard’s polar security vutter (PSC) initiative to reconstitute the service’s heavy icebreaker fleet.
"The lead ship in the planned three-vessel fleet may not arrive until 2033, representing a nine-year delay from the coast guard’s original delivery goal, according to the watchdog. Additionally, the program’s current baseline shows the first cutter’s procurement, construction and improvements costs will total US$3.42 billion — nearly the cost for all three boats under the initial baseline.”
Bollinger has vigorously denied that the first vessel will be delayed until 2033, saying that this date mentioned in the GAO assessment is the program’s worst-case, “threshold date,” and includes substantial risk margin. Given the six-year delays foreseen so far, who could reasonably expect further delays to the PSC programme (that’s a rhetorical question; please don’t answer it in an email to me)?
The shipbuilder did not deny, however, that the first vessel would now likely cost over US$3 billion, a mind-blowing sum for a single icebreaker, and nobody has yet explained to me why three such heavy-duty polar ships are operationally necessary for the US.
The PSCs will each have accommodation for up to 186 crew, scientists and other personnel, and an endurance of 90 days. Given the axing of the budgets for oceanographic science, it is not clear where those scientists would come from, but that is likely a problem for the 2030s, along with whether multi-year polar sea ice 1.8 metres thick will still be an issue ten years hence in our warming world.
We don’t want to cast aspersions on the previous management of Halter Marine, which was originally awarded the contract by the coast guard in 2019 for the PSC construction, but we were surprised to learn that Bollinger had had to buy a specialty press to bend the EQ47 high-strength, low-temperature steel needed for the hulls of the PSCs. What kind of shipyard wins a contract to build a series of hulls but doesn’t possess the press needed to shape the blocks for five more years?
Meanwhile, steel has already been cut on the six smaller Arctic security cutters (ASCs), icebreakers that are considerably cheaper than the larger and more powerful PSCs. Bollinger’s Houma yard in Louisiana was awarded US$2.2 billion for four hulls under this ASC programme, and Rauma in Finland was awarded US$1.1 billion for two hulls, meaning that six smaller, 10,100kW icebreakers will cost less than one PSC. All six ASCs are supposed to be delivered in 2031 from their 2026 order date, likely ahead of the first PSC, and it will be interesting to see whether the Finns can beat the Americans in delivering the first ship.
These ASCs are designed to break one metre of ice at four knots. The ASCs were designed in Vancouver by Seaspan, as per the US Naval Institute, whereas the PSCs were ordered to no fixed design. The ASC uses a hull shape developed by Aker Arctic Technology, which is the same as the multi-purpose icebreaker design developed for the Canadian Coast Guard, and the ships each have a total complement of 85.
Who knew that in shipbuilding you get economies of scale from building more identical ships, and that building to a proven design saves time and cost? Again, it is not necessary for readers outside Washington, DC, to answer that question.
As Bollinger tools up, another company has entered the naval shipbuilding business in the US.
Anduril Industries has been adept at both managing investor expectations and the US defence procurement system through impeccable political connections in both Washington and Canberra.
Anduril Industries is privately held and was founded in 2017 by Palmer Luckey, Brian Schimpf (the current CEO), Trae Stephens (now Executive Chairman), Matt Grimm, and Joe Chen. Hawaiian shirt-wearing Mr Luckey is the “larger than life” public face of the company. He made his fortune by selling the Oculus virtual reality glasses technology business to Meta (then Facebook) for around US$2 billion in 2014 (this gaming and entertainment technology is not to be confused with Meta’s notorious smart eyewear products, which are colloquially also known as "pervert glasses").
Brian Schimpf and Trae Stephens came from Palantir Technologies, the data analytics company focused on defence and intelligence contracts, founded by right-wing billionaire Peter Thiel, who recently moved to Argentina and has invested in shale oil there. Palantir has been repeatedly attacked for some of its less ethical defence initiatives but has won major American military contracts, including the so-called Golden Dome missile shield project.
Big name venture capital companies have supported Anduril; these include Andreessen Horowitz, Thrive Capital, Founders Fund, and Lux Capital. With the company's rising value, the investors and founder have made out like bandits.
Anduril has raised over US$10 billion over multiple rounds of investor funding. The company reached a US$61 billion valuation in May 2026, making lucky Mr Luckey and his colleagues billionaires.
This surge in valuation came after the US Army awarded the company a contract worth up to US$20 billion in March 2026 covering, “AI-driven software systems for integrating autonomous sensors and effectors across a battlespace – along with Ghost autonomous aerial systems, the Roadrunner reusable jet-powered loitering munition, and the Pulsar counter-drone electronic warfare system.”
Anduril is one of the most valuable defence technology companies ever built by a private startup… and if you thought the bill for the Aukus nuclear submarine programme was expensive, don’t worry Mr Luckey is keen to get his share... Aussie taxpayers are already helping make American billionaires even richer.
The fact that Mr Luckey’s company has never built a piece of a nuclear submarine before and will be building his facility on the site of a long-closed Bethlehem steel shipyard is not a problem. He is also bidding to acquire a Nissan car factory in Japan to turn into a drone factory.
Despite its relatively recent foundation, Anduril is already a contractor the Royal Australian Navy (RAN). As we reported last year, the RAN awarded the company a AU$1.7 billion contract to Anduril’s Australian subsidiary to deliver a large fleet of autonomous underwater vehicles (AUVs) which it has branded as Ghost Sharks over the next five years.
The specifications have not been released, but the AUVs will be all-electric and will be less than 12 metres long. Anduril announced the successful program of record designation in a press release last year after saying it successfully completed the co-development contract with the Australian Department of Defence (DOD), and delivered three Ghost Shark extra large AUVs ahead of schedule and on-budget.
An additional AUV of the same design was also delivered to the US DOD. The prototypes were the culmination of an AU$140 million (US$98 million) co-development contract to design and develop three AUVs in three years.
The former offshore support vessel Normand Jarl (now ADV Guidance) was acquired by the RAN from Solstad in Norway in April 2023 at a price of AU$110 million (US$77 million) to act as the mothership for these AUVs, we understand. Effectively, the RAN has been using the same methodology as Ocean Infinity, which mastered the concept with the subsea vessel Island Pride, but Anduril will deploy them for military use rather than searching for wrecks with Kongsberg Hugin AUVs as the British company did.
However, we note now that Ocean Infinity also has certain of its class of “lean crewed” survey vessels on charter to the Royal Navy in the UK, so it too has pivoted to a defence niche, following the money. If Ocean Infinity were to be sold, as has been rumoured at various times, I would wager that Anduril would be on the shortlist of buyers.
The new 7,400m² Anduril facility in Sydney is purpose-built to produce Ghost Shark AUVs, and also to manufacture its commercial equivalent, an AUV that the company has called the Dive-XL, as it goes head to head with Kongsberg’s Hugin series, at-scale. So far, no release has been issued on the delivery of the next wave of AUVs from the facility.
However, this month Anduril announced it was doubling down in the naval and shipbuilding arena.
Anduril has announced that it plans to spend US$3.7 billion to construct what the company calls a “software-defined shipyard” at Sparrows Point in Baltimore County on the US East Coast. The company won a US Navy contract worth as much as US$2.9 billion that is contingent on “demonstrated production outcomes.”
How hard can shipbuilding be compared to making virtual reality headsets?
Anduril has named the facility “Arsenal-2,” (presumably not after the English football club; its first factory for drones and missiles in Ohio is called Arsenal-1). It said that the new 187-acre (76-hectare) site (“more than 40 football fields”) will employ 3,100 people, adding nine million labour hours each year, or a 15 per cent increase, for the Virginia-class nuclear submarine program, to which Anduril will manufacture subcomponents including torpedo launch tubes. It will commence initial production in Orange Country, California, as the Baltimore facility is constructed and fitted out, and staff are hired there.
It is worth quoting Anduril’s statement on what will make Arsenal-2 better and different from all existing shipyards. Bollinger’s woes have shown the problems of shipbuilding in the US. Anduril hopes to overcome these by applying the power of Silicon Valley software:
“ArsenalOS, Anduril's industrial software platform, will provide the digital backbone for the yard and its production systems. It will connect fabrication workflows, outfitting sequences, material movement, inspection protocols, and documentation requirements in a single system, giving workers and managers real-time visibility into the state of production. Digital work instructions, embedded quality controls, modern sensing, and automation will support the expertise of welders, pipefitters, electricians, shipfitters, inspectors, and other skilled trades. By making production requirements more visible and processes more consistent, ArsenalOS will help new workers train faster while preserving the technical rigour, judgment, and quality required for submarine construction.”
That’s what Bollinger had been missing in Mississippi, clearly.
Anduril and Ocean Infinity bring together the new wave of investment in dual purpose technology, both defence-focused, but also with survey, renewables, salvage and oil and gas applications. Notably, however, Ocean Infinity has never attempted to build its own crewed or uncrewed vessels, preferring to be an integrator of ships built by Vard in Vietnam, together with AUVs designed and built in Norway.
In July, we highlighted that Thales, Fincantieri and Lockheed Martin were all buying into offshore and subsea technology companies. The wars in the Persian Gulf and in Ukraine have stimulated a wave of investment in dual-use maritime technology of a size never before seen in the industry.
With the support of Pete Hegseth and the Pentagon, Anduril is winning billions of dollars of contracts and has a valuation ten times larger than Fincantieri, the owner of Vard and one of Europe’s largest naval shipbuilders. Anduril’s valuation is now similar to that of BAE Systems, the British submarine and weapons giant.
However, shipbuilding is a shocking-difficult business to execute consistently profitably, unless you have a customer with an open chequebook willing to pay for all the problems encountered, as Bollinger has had for the “troubled” PSC programme. Will taxpayers continue to fund cost overruns and delays? I suspect increasingly in an era of rising defence budgets and rising budget deficits across the developed world, but in the US especially, something will have to give.
Whilst Anduril is riding a wave now, it remains to be seen whether its digital ambitions and AI in shipbuilding will match the cruel realities of bending steel in a covered yard and welding blocks together cheaply and quickly to specification.
Shipbuilding is fundamentally a physical business producing a physical product. We wish Palmer Luckey luck in his efforts to transform an industry that has been a graving dock graveyard of investor dreams for several centuries.
