Anduril Industries: A Full Guide to Lattice, Fury, the Arsenal Factories, Steelfin, and Honest Limits

On October 6, 2026, Anduril Industries put $3.7 billion of its own capital behind a submarine shipyard in Baltimore County, matched by a Navy contract worth up to $2.9 billion. Six days later it unveiled Steelfin, a mass-producible family of surface-to-air interceptors. A company that started in 2017 selling sensor towers and software is now bidding to build fighters, missiles, and submarine sections.

That's the tension this guide sits on. Anduril's 2025 revenue doubled to $2.2 billion and its May 2026 Series H valued it at $61 billion. Its products have also failed in public tests and on the Ukrainian front, according to Wall Street Journal reporting.

Here's the shape of the piece: what Anduril Industries actually sells, how Lattice ties it together, where Fury, the Arsenal factories, Steelfin, Ghost Shark, and EagleEye stand as of October 2026, how the money works, and the honest limits buyers and investors should price in. For the wider field, keep Inside Deep Tech's Anduril competitors breakdown open in another tab.

Last updated: October 12, 2026.

Key Takeaways

• Anduril reported **$2.2 billion** in 2025 revenue and raised **$5 billion** at a **$61 billion** valuation in May 2026.
• Lattice is the core product: the Army picked it for NGC2 in a deal worth up to **$1.8 billion** over five years.
• Fury (FQ-44A) won a CCA production contract in June 2026, and the first Ohio-built Fury rolled out July 27, 2026.
• New bets: the **$6.6 billion** Arsenal-2 submarine yard and Steelfin interceptors, flight testing in 2027.
• Honest limit: drone boats, Fury, Anvil, and Altius all hit documented setbacks, and Arsenal-2 opens in 2030.

What Anduril Industries actually is

Anduril was founded in 2017 by Palmer Luckey, the Oculus founder, with a thesis that the Pentagon would buy autonomous hardware the way tech companies buy products: off the shelf, iterated fast, and built on company money rather than cost-plus development contracts. CEO Brian Schimpf framed the Series H letter around defense industrialization and resilient production capacity.

The business has three layers. Software comes first: Lattice, the command-and-control and autonomy platform. Then hardware: aircraft, interceptors, undersea vehicles, counter-drone systems, and soldier wearables. Then factories: the Arsenal network that turns those designs into volume.

That last layer is the newest and the most expensive. It's also where Anduril now looks least like a startup and most like Lockheed Martin or General Dynamics.

Funding, valuation, and revenue

In May 2026 Anduril closed a **$5 billion** Series H at a **$61 billion** valuation, led by Thrive Capital and Andreessen Horowitz. That was more than double the **$30.5 billion** valuation from its **$2.5 billion** June 2025 round led by Founders Fund, and it lifted total funding past **$11 billion**.

Revenue kept pace on paper: **$2.2 billion** in 2025, roughly double the year before. In July 2026, TechCrunch relayed Reuters reporting that Anduril was in talks at about a **$100 billion** valuation. Anduril's response was that no decisions had been made, so treat that figure as unconfirmed.

For context on where that capital sits among peers, see Inside Deep Tech's look at the defense tech startups attracting record funding in 2026.

Lattice: the software layer everything plugs into

Lattice is Anduril's operating system for sensors, weapons, and autonomous vehicles. Early demos connected sensors and weapons built by different companies into one battle network. Today Anduril runs it across its own aircraft, counter-drone systems, and undersea vehicles, and it now underpins the Army's command-and-control modernization.

The Army's Next-Generation Command and Control (NGC2) program is the biggest proof point. In June 2026 the Army set up the NGC2 common data baseline and picked Anduril to lead it. On October 5, 2026 Anduril announced a follow-on with a **$162.8 million** base period and options worth up to **$1.8 billion** over five years, starting with the Army's First Corps, which commands four divisions across the Pacific.

The field numbers are what got the Army's attention. At the Ivy Mass exercise, more than **2,500** end-user devices ran on the same data architecture, and NGC2 cut artillery fire timelines by **90%** versus legacy systems, according to Anduril.

Here's why that matters for you if you build sensors or effectors: once Lattice is the data layer, integrating with it stops being optional. Anduril says no single company should control the whole command-and-control stack. Rivals will hold it to that.

Fury (FQ-44A): the Collaborative Combat Aircraft

Fury is Anduril's uncrewed fighter for the Air Force's Collaborative Combat Aircraft (CCA) program. It went from clean-sheet design to first flight in 556 days, flying semi-autonomously from the start rather than with a remote pilot. That first flight came on October 31, 2025.

On June 17, 2026, the Air Force awarded engineering, manufacturing development, and production contracts to both General Atomics (FQ-42A) and Anduril (FQ-44A). The goal is at least **150** CCAs combined by the end of the decade, toward an eventual fleet of around **1,000** CCAs with a combat radius requirement of at least **700** nautical miles for Increment 1.

The contract value is classified. The Air Force said it is meeting its goal of one-third the cost of an F-35, roughly less than **$30 million** per aircraft, and that lot sizes and unit costs will be released later.

One detail matters more than the headline. The Air Force split airframe from autonomy. Anduril, Shield AI, and Collins Aerospace each got six-month options to mature mission autonomy software, with a downselect to an initial provider by summer 2027. Winning the airframe doesn't guarantee Lattice flies it.

Arsenal-1: the Ohio factory behind the pitch

Arsenal-1 sits near Rickenbacker International Airport in Pickaway County, outside Columbus. Anduril began Fury production there in March 2026, three months ahead of its July target, and the first Ohio-built Fury rolled off the line on July 27, 2026, 557 days after the project was announced.

The line can build up to **150** Fury aircraft a year in its current configuration. At full buildout the campus is planned at five million square feet on **500** acres, with **$910.5 million** in committed capital investment and **4,008** jobs by 2035.

Of course, a capacity number is not a delivery number. The Air Force hasn't published Fury lot quantities, so the honest read is this: the factory exists and produced its first jet on schedule. The production ramp is still unproven.

Anduril product lines at a glance

Source: Anduril NGC2; Air & Space Forces Magazine; Columbus Region; Anduril Arsenal-2; Anduril Steelfin; Janes; DefenseScoop. Status as reported on each date.

Arsenal-2: Anduril moves into submarines

Arsenal-2 is Anduril's first entry into a major legacy program. The company is putting in **$3.7 billion** of private capital, and the Navy awarded a contract worth up to **$2.9 billion**, for a combined **$6.6 billion** to build Virginia-class submarine components and large assemblies for General Dynamics Electric Boat and HII's Newport News Shipbuilding.

The site is Tradepoint Atlantic at Sparrows Point in Baltimore County, Maryland. Anduril plans more than two million square feet on a **187**-acre site, a **1,200**-foot graving dock, **3,100** direct jobs, and more than **11,000** indirect jobs. Initial operations are expected in 2030.

Two details deserve attention. First, Navy payments are tied directly to demonstrated production outcomes, so Anduril carries most of the execution risk. Second, before Arsenal-2 opens, a **160,000**-square-foot Orange County facility will prototype torpedo tubes and work through SUBSAFE qualification.

Scale check: this summer General Dynamics won a **$76.6 billion** Navy award to build submarines. Anduril is adding a supplier lane to that enterprise, not replacing it.

Steelfin: interceptors built for volume

Unveiled on October 12, 2026, Steelfin is a two-stage family of surface-to-air interceptors aimed at cruise, ballistic, and hypersonic missile threats. The first stage swaps solid rocket boosters for different ranges; the second stage keeps a common terminal interceptor with shared airframe, avionics, and software.

The pitch is the cost-exchange problem. Air defense units keep firing expensive interceptors at cheap threats. Anduril says it plans on the order of several thousand Steelfin rounds per year in the short term, compatible with the M903A2 launcher, the Mk 41 Vertical Launch System, Typhon, and fire control such as AEGIS and IBCS.

At first glance that reads like a finished product. It isn't. Anduril has completed ballistic testing with representative air vehicles and starts flight testing next year. No published unit price, no flight data, no program of record yet.

Ghost Shark and EagleEye: the allied and soldier bets

Ghost Shark is an extra-large autonomous undersea vehicle built for the Royal Australian Navy from Anduril's Dive-XL. It grew out of a three-year, US$100 million co-funded research program that delivered three prototypes in 2025. Anduril Australia now holds an A$1.7 billion (about US$1.1 billion) program of record, with a Sydney factory that opened in October 2025 and first delivery planned for January 2026.

EagleEye is Anduril's heads-up display for the Army's Soldier Borne Mission Command (SBMC) program, the successor to Microsoft's troubled IVAS headset that Anduril took over in 2025. The company plans **170** phase-two prototypes, with testing in early 2027 and a possible production downselect in fiscal 2028. The stakes are large: IVAS was expected to be worth up to **$22 billion** before it was scrapped.

Both programs show the same pattern. Anduril wins prototype and early production work fast. Full-rate production decisions come years later.

Anduril vs the primes and the new entrants

Anduril now competes on two fronts. Against the primes (Lockheed Martin, Northrop Grumman, General Dynamics, RTX), it argues speed and self-funded development. Against new entrants, it argues breadth and factories.

The new entrants are well funded too. TechCrunch notes Shield AI raised **$1.5 billion** at a **$12.7 billion** valuation in March 2026, and Helsing was reportedly near a round at about **$18 billion**. Shield AI is already competing with Anduril on CCA autonomy.

Top 10 Anduril Competitors in the Modern Defense-Tech Industry

Who competes with Anduril across autonomy, drones, counter-UAS, and defense software, and where each rival is strongest.

Inside Deep Tech

For the drone-specific side of the market, Inside Deep Tech's guide to autonomous drones and AI defense in 2026 covers the dual-use players Anduril meets in counter-UAS and loitering munitions.

Honest limits

• Test and combat setbacks are documented. WSJ reporting relayed by TechCrunch cites more than a dozen drone boats failing in a May 2025 Navy exercise, a Fury engine damaged in a ground test, and a **22**-acre Oregon fire from an Anvil test.
• Ukraine was rough. Ukraine's SBU reportedly stopped using Altius loitering drones in 2024 after crashes and jamming problems. Anduril says the issues are typical of weapons development.
• Valuation runs far ahead of revenue. **$61 billion** on **$2.2 billion** of 2025 revenue is roughly 28x sales for a private company with no public audited accounts.
• Big numbers are mostly future numbers. Arsenal-2 opens in 2030 with outcome-tied payments, Steelfin has not flown, and the NGC2 ceiling depends on options beyond a **$162.8 million** base.
• Political proximity cuts both ways. Co-founder Palmer Luckey joined the Pentagon's Project Meridian days before the Navy award, which invites scrutiny of how contracts get awarded.

Inside Deep Tech's take

Inside Deep Tech's take: Anduril's strongest asset is not Fury or Lattice alone. It's the willingness to spend its own money on factories before the contracts arrive, then accept payment terms tied to output. That's a real break from cost-plus defense contracting.

The weakness is the same move seen from the other side. Every new Arsenal adds fixed cost to a company whose public record still includes failed exercises. If CCA lots, NGC2 options, and Steelfin flight tests land in 2027, the valuation looks earned. If two of the three slip, it doesn't.

FAQ

What to watch next

Three dates will tell you more than any funding round. Summer 2027, when the Air Force picks an initial CCA mission autonomy provider. 2027, when Steelfin flight testing begins. And 2030, when Arsenal-2 is due to open with payments tied to output.

Between now and then, track Fury lot quantities, NGC2 options exercised beyond the base period, and the first Ghost Shark fleet numbers Australia chooses to publish. Anduril Industries has proven it can build factories fast. The next two years test whether it can fill them.