Anduril Built a Submarine Factory 18 Months Before It Had a Contract

Why Anduril Chose Sydney for Its First Overseas Submarine Factory

The 7,400-square-metre facility that opened in Sydney on October 31, 2025, wasn't a response to a contract. It was a bet placed before one existed. Anduril's Asia Pacific CEO David Goodrich asked his board to fund the factory 18 months before the Royal Australian Navy awarded the Program of Record for the Ghost Shark Extra Large Autonomous Underwater Vehicle. "I had a lot of sleepless nights," Goodrich told Forbes Australia. "We had to build it. Otherwise there was no way we could get into full production in the time we said we would."

That timeline turned out to be almost absurdly compressed. Just 51 days after the contract was announced in September 2025, the first Ghost Shark rolled off the Sydney production line ahead of schedule, ready for in-water testing ahead of planned delivery to the Royal Australian Navy in January 2026. The factory was already operating before the deal was signed.

The strategic logic runs deeper than one contract. Anduril had spent the prior three years executing a co-development agreement to design and build three Ghost Shark XL-AUVs, and delivered all three ahead of schedule and on budget. That track record gave the Australian government enough confidence to commit to a full fleet delivered over five years. The Sydney factory is purpose-built not only for the military variant but also for the commercial Dive-XL baseline, with export to allies flagged as a future revenue stream pending government approval.

Australia also offered something Anduril couldn't easily replicate in the U.S.: a government willing to move at a company-matching pace. Defence Industry Minister Pat Conroy attended the factory opening and called the Ghost Shark "the most high-tech long range autonomous underwater capability that exists in the world today." The facility has already drawn on a supply chain of more than 40 Australian SMEs and created over 150 high-skilled local jobs.

The Sydney buildout fits a pattern Anduril founder Palmer Luckey has described explicitly: design weapons that can be mass-produced in existing industrial facilities. "We're trying to build submarines that can be built in car factories," Luckey told Forbes Australia. The Ghost Shark program uses industrial robotic manufacturing and AI-driven logistics, equipment and processes that can, in theory, be replicated across Australia's broader industrial base if production needs to scale tenfold.

Anduril's broader financial position gave it the runway to make that bet. In June 2025, the company closed a Series G round led by Founders Fund — the largest check the firm has ever written. That round valued Anduril at roughly double its valuation from six months earlier, after the company doubled its annual revenue in 2024. The same month, Anduril formally took over the US Army's Integrated Visual Augmentation System from Microsoft, a contract that cemented its status as something closer to a prime defense contractor than a startup.

Sydney is Anduril's first overseas submarine factory, but it isn't the only manufacturing buildout underway. The company is constructing Arsenal-1, a hyperscale manufacturing campus in Pickaway County, Ohio, modeled on Tesla's Gigafactory concept, set to open in 2026 and eventually employ 4,000 additional workers. It also operates a high-volume solid rocket motor plant in Mississippi. The Australian facility is smaller (deliberately so) and designed to prove that autonomous undersea vehicles can be manufactured outside the U.S. defense-industrial base, by a distributed workforce, at a pace that legacy primes can't match.

What the Roles Reveal About Autonomous-Undersea Talent

Anduril's Australian hiring pipeline is small but specific, and the roles tell you exactly what building autonomous undersea systems actually requires. Sydney-based positions span electronics test, systems engineering, and field operations. Strip away the job-posting boilerplate and a coherent picture emerges: Anduril isn't looking for generalists. It wants engineers who can move between hardware and software, between a factory floor and a naval deployment, without waiting for someone to hand them a spec sheet.

The most instructive listing is the Senior Systems Engineer role in Sydney. The description ties directly to the XL-AUV program, the same effort underpinning Anduril's contract with the Australian Defence Force. The job asks for someone who can "own technology solutions deployed to customers," act as a subject matter expert across hardware selection, testing, robotic perception, and cybersecurity, and "work holistically on complex robotic systems across design, implementation, operation and sustainment." That last phrase is doing real work. Anduril doesn't want a design engineer who hands off to a test team. It wants one person who follows the system from architecture through sea trials and into the maintenance cycle.

The required qualifications reinforce this. A bachelor's degree in robotics, mechatronics, computer science, or engineering (or "equivalent experience," a door Anduril leaves deliberately wide). The capacity to "act as the technical owner for a complex subsystem" including stakeholder engagement, requirements definition, roadmap management, and team coordination signals a role that blends program management with deep technical work. Preferred qualifications add experience with "robotic and/or autonomous systems" and familiarity with "the defence, maritime and/or aerospace domains," but neither is mandatory. Anduril is willing to train domain expertise if the systems-engineering foundation is solid.

The Naval Architect listing in Sydney is narrower but equally revealing. Posted at the mid-senior level, it's a traditional marine-engineering role: hull design, stability analysis, structural integration, sitting inside a company whose entire pitch is software-defined autonomy. The pairing tells you something about where autonomous undersea vehicles still depend on conventional naval architecture: the hull still has to survive pressure, corrosion, and hydrodynamic loads before any AI onboard matters. Anduril needs people who understand both halves.

Then there's the Field Operations Engineer for Autonomous Underwater Vehicles, a role that exists almost nowhere else in the defense industry. This person writes and executes test plans for sea trials, monitors AUV sensor data and mission progress in real time, coordinates launch-and-recovery evolutions, performs field maintenance and troubleshooting, and serves as the technical liaison between Anduril's internal engineering teams and defense customers. The posting calls for a B.S. in software or mechanical engineering, five years of multidisciplinary experience, and, critically, the "ability to drive challenging and vague technical problems to clarity and resolution" in unstructured situations at the hardware-software-networking interface. This is not a desk job. The posting notes up to 20% travel to co-locate with end users, and the role requires an NV2 security clearance.

What these three roles share is a refusal to separate the digital from the physical. The systems engineer works across Lattice OS, Anduril's AI-powered command-and-control platform, and the vehicle's hardware. The naval architect designs structures that will carry autonomous payloads. The field engineer troubleshoots both in open water. Anduril's Sydney hiring isn't building a software team that happens to work on boats. It's building an engineering workforce that treats the ocean as an operating environment for robotic systems, and that demands fluency in both.

For engineers weighing whether to move into defense, the signal is clear: the hardest problems in autonomous undersea systems aren't in any single discipline. They're at the intersections — where autonomy meets hydrodynamics, where sensor fusion meets saltwater corrosion, where a mission plan meets a failed actuator at depth. That's what Anduril is hiring for.

How a Dual-Continent Manufacturing Strategy Reshapes Defense Hiring

Anduril's Sydney submarine factory didn't appear in a vacuum. It's the overseas counterpart to a domestic manufacturing buildout so large it has reshaped the company's entire hiring trajectory, and the comparison between the two sites reveals how defense-AI talent is being distributed across continents.

Start with the scale of Arsenal-1. Anduril's flagship production facility in Pickaway County, Ohio, spans more than 5 million square feet and is expected to generate over 4,000 jobs. Production began ramping in March 2026. The plant is designed around what Anduril calls "hyperscale" manufacturing: the idea that autonomous defense systems can be built at a volume and cost point that resembles commercial aerospace more than traditional defense procurement. That philosophy demands a specific workforce: engineers and technicians fluent in robotic assembly, AI-driven quality systems, and modular hardware design, hired in bulk.

Sydney operates on a different logic. The Ghost Shark facility is purpose-built for a single product line: the XL-AUV autonomous undersea vehicles the Royal Australian Navy ordered under the Program of Record. The factory combines robotic manufacturing with AI-driven logistics and an on-site test tank for in-water verification before sea trials. It's smaller, more specialized, and tied directly to a sovereign defense requirement: Australia wants undersea drones built on Australian soil by Australian workers.

The hiring profiles reflect that divergence. Arsenal-1's 4,000-plus roles skew toward high-volume production engineering, supply chain management, and manufacturing operations at scale. Sydney's demand is narrower and deeper — autonomous systems engineers, undersea vehicle integration specialists, and technicians who can work at the intersection of naval architecture and machine learning. Zero G Talent's board currently lists 134 active Anduril roles added in the past week alone, with positions like Electronics Test Technician in Sydney sitting alongside supply chain and recruiting roles in the U.S., a snapshot of the company's simultaneous dual-continent staffing push.

Chris Brose, Anduril's Chief Strategy Officer, signaled this approach as far back as August 2024, saying the company was considering basing future Arsenal-style plants overseas in allied nations. The May 2026 Series H round made that explicit, with funding targeted at scaling production capacity at Arsenal-1 and "similar facilities." Australia is the proof of concept. Japan, where Anduril has explored a facility called "Arsenal J" since roughly December 2025, may be next.

What this means for defense hiring is structural, not incidental. Anduril isn't just opening factories abroad; it's building parallel talent ecosystems. Ohio feeds the mass-production backbone. Sydney feeds sovereign allied capability in the Indo-Pacific. Engineers choosing between the two aren't just picking a city — they're picking a mission profile. And the company's pace suggests both ecosystems will keep expanding. The first Ghost Shark production variant rolled off the Sydney line ahead of schedule, with full-scale production slated for 2026. Arsenal-1 is still ramping. For anyone tracking where defense-AI hardware jobs are heading, the answer is no longer a single zip code. It's two continents, two factory floors, and a hiring pipeline that runs through both.

Can the Silicon Valley Ethos Coexist with the ADF?

Anduril's Sydney factory sits next to a Bunnings hardware store in a suburban industrial park. That placement is deliberate, and it tells you almost everything about the culture clash unfolding inside the company's Australian operation.

The company's founding playbook came out of Palmer Luckey's experience building Oculus Rift and Trae Stephens' frustration with Palantir-era procurement cycles. Anduril was designed to operate like a venture-backed startup that happens to sell to militaries: develop products before the Pentagon asks for them, iterate fast, and use commercial technology as the backbone. Brian Schimpf, Anduril's CEO, has described the model as "scaling a lot of 300- to 500-person companies" rather than one centralized 6,000-person bureaucracy. That structure gives teams autonomy. It also assumes a workforce comfortable with ambiguity, rapid pivots, and a pace that would be unusual inside a traditional defence prime.

Australia's defence-industrial base has operated on a different clock. Programs of record, the formal process by which the Australian Defence Force acquires major capabilities, have historically moved in multi-year cycles with extensive documentation, staged gates, and heavy oversight. The relationship between contractor and client is structured, hierarchical, and risk-averse. Anduril's counter-UAS trial at RAAF Base Darwin illustrates the tension. Since October 2024, the company has delivered four hardware and sixteen software updates at no additional cost to the Royal Australian Air Force. David Goodrich, executive chairman and CEO of Anduril Australia, noted that the trial still doesn't include effectors to actually intercept threats, not because the technology isn't ready, but because Australian law doesn't yet permit it. "The Australian government doesn't give us the opportunity to fly effectors out against these threats," Goodrich said at the Avalon airshow. "By the way, it's the same in the United States and it's the same all around the world. We need to fix that."

Air Vice-Marshal Nicholas Hogan, head of air force capability for the RAAF, acknowledged the gap from the other side: "In order to use the full range of capabilities we would like to use in a domestic environment, we will require some legislative changes." He added that legislation moves slower than the technology it's trying to regulate.

That mismatch — Anduril's software-speed iteration against the ADF's procurement-speed approval — is the central friction. But it's also where the opportunity sits. The Ghost Shark program compressed what would normally be a decade-long acquisition into roughly three years from contract to production vehicle. As Patrick Tucker reported in Defense One, Australia got a new sub drone "far faster than the US Navy could have."

The workforce implications are specific. Anduril's Sydney operation needs engineers who can work inside Lattice, the company's AI command-and-control platform, and also navigate the ADF's certification and safety requirements. They need software developers comfortable pushing updates every two weeks alongside defence officers who are accustomed to reviewing changes over months.

Goodrich, who has three decades of international corporate experience including a decade advising the ADF, was likely chosen to lead the Australian entity precisely because he can translate between those worlds. Anduril established it as an independent entity, Anduril Australia, rather than a branch office, a structural choice that gives the local team room to adapt the Silicon Valley playbook to Australian regulatory and cultural conditions without running every decision through Costa Mesa.

The company has also moved to embed itself in the local supply chain. More than 40 Australian companies are already part of the Ghost Shark supply chain, with that number expected to grow. That localisation matters politically — the Albanese government has made sovereign capability a defence-industrial priority — and it matters practically. Anduril's design philosophy, which Luckey describes as building "submarines that can be built in car factories" and "cruise missiles that can be built in farm implement factories," depends on being able to source from the industrial base that already exists in-country.

The cultural integration is incomplete. Anduril's Silicon Valley ethos — Luckey in Hawaiian shirts and shorts giving lobby tours, the company's first acquisition being a gaming engine, the open embrace of military work that much of the tech industry still avoids — sits uneasily alongside the ADF's institutional conservatism. But the Ghost Shark factory is open, the first vehicle is in sea trials, and the RAAF is training operators on Lattice. The friction is real. So far, it's productive.

What Anduril's Australia Bet Signals for Allied Defense-Tech Workforces

The AUKUS partnership is quietly rewriting how allied nations build weapons, and who builds them. At the May 2026 Shangri-La Dialogue in Singapore, the defense ministers of Australia, the United Kingdom, and the United States announced the first Pillar II "Signature Project": joint development of payloads and enabling systems for uncrewed undersea vehicles, with delivery starting in 2027. That single line in a joint statement carries enormous implications for where defense engineers will work over the next decade, and what "allied interoperability" will actually mean on a factory floor.

The backdrop is the U.S. Navy's own submarine crisis. The Congressional Research Service reported in July 2025 that the service's next-generation attack submarine, SSN(X), has been pushed from the 2030s into the 2040s due to budget constraints, a delay the Navy's own fiscal 2025 budget proposal called a "significant challenge to the submarine design industrial base." The Columbia-class ballistic missile submarine is already projected to arrive two years late, in 2029. The GAO has described U.S. shipbuilding as being in a "perpetual state of triage."

Australia's answer has been to skip a generation. Rather than wait for Washington to sort out its submarine pipeline, Canberra signed a contract with Anduril in September 2025 for dozens of Ghost Shark XL-AUVs, with the first entering service in January 2026. The program went from prototype to fleet contract in three years, a speed that would be unrecognizable inside the Pentagon's acquisition system.

That speed is precisely why Anduril's Sydney workforce matters beyond Australia. The May AUKUS ministers' statement confirmed that the three nations are collaborating on common payloads for uncrewed underwater vehicles, building on exercises like "Autonomous Warrior" under Pillar II's Maritime Big Play Initiative. The U.S. is drawing on the Ghost Shark platform itself; the UK is developing a parallel capability through Project Cetus and the Excalibur UUV. A shared operational infrastructure — common command-and-control systems, interchangeable payloads — is the logical next step. When that arrives, an engineer in Sydney working on Ghost Shark autonomy will be building something a Royal Navy operator in Portsmouth can plug into.

This is the distributed allied defense-production model taking shape in real time: not one country designing and others buying, but multiple sovereign workforces building interoperable systems side by side. The AUKUS license-free environment for defense trade is expanding, with the ministers committing to "narrow the list of excluded technologies," a direct response to the information-sharing barriers that have hampered the partnership.

For the talent market, the signal is concrete. Anduril's Sydney operation is hiring electronics test technicians, systems engineers, and supply-chain specialists, the kind of roles that anchor a sovereign manufacturing base. These aren't outsourced support positions. They're the workforce for a factory that is simultaneously an Australian sovereign capability and a node in a three-nation autonomous-systems network.

The U.S. submarine industrial base will eventually recover. SSN(X) will get funded, the design will mature, boats will be built. But in the interim, and that interim now stretches across the 2030s, the center of gravity for autonomous undersea systems development has shifted. It sits in a facility in Sydney where engineers are building drones that three navies intend to operate as a fleet.

Working in frontier tech? Zero G Talent tracks the openings: browse frontier tech jobs, openings at Anduril Industries, and the people building the field.