Anduril: Resurrecting the Arsenal of Democracy

"We must be the great arsenal of democracy. For us this is an emergency as serious as war itself. We must apply ourselves to our task with the same resolution, the same sense of urgency, the same spirit of patriotism and sacrifice as we would show were we at war."
— Franklin D. Roosevelt, Fireside Chat, December 29, 1940

In July of this year, this publication authored The Profiteer State, an accounting of what has become of American defense manufacturing. The numbers in that piece were not in dispute and remain undisputed. 54% of the Pentagon's $4.4 trillion in discretionary spending between 2020 and 2024 flowed to private firms, with nearly $800 billion of it going to five companies. Lockheed Martin alone took $313 billion. Those same firms spend north of $80 million a year on lobbying and have moved nearly $280 million into political campaigns since 1990, and they do it because every dollar of it comes back tenfold.

We also said something else in that piece: We said the problem is structural. Under cost-plus contracting, where margin is calculated as a percentage of cost, a firm that runs expensive and slow is not cheating the system. It is, in fact, reading the incentives correctly. So, how can this problem be fixed? Simple: Change the incentive. Fourteen months ago, Anduril Industries ran that test.

Arsenal-1

On January 16, 2025, Anduril Industries announced it would build a hyperscale weapons factory called Arsenal-1 in central Ohio, taking over an existing 700,000-square-foot structure with room to grow to five million, and that the first products would come off the line in July 2026. The company committed roughly a billion dollars of its own capital to the project. Defense announcements of this kind are made constantly and mean very little. The industry's own term for what usually happens next is the "valley of death," the gap between a working prototype and anything resembling mass production. Production lines went hot at Arsenal-1 in March 2026, three months early. The company's co-founder and chief operating officer, Matthew Grimm, noted at the time that the project was running both ahead of schedule and under budget, and had the good sense to acknowledge that this is a rarity in the defense business. On July 27, 2026, the first Ohio-built Fury rolled off the line. From groundbreaking announcement to finished combat aircraft: 557 days. At full build-out the site is projected to employ more than 4,000 Ohioans by 2035, making it the largest single job-creation project in the state's history. Set that against the F-35 program, launched in 2001. Full-rate production was not formally approved until 2024, 23 years later, and a program whose lifetime cost estimates now exceed two trillion dollars. We are not comparing apples to oranges here. The distinction is not merely between a drone and a fifth-generation stealth fighter. The observable distinction is between two institutional metabolisms, and the gap between them is an order of magnitude.

The Model Inverted

The mechanism behind the speed is no secret. Under the cost-plus arrangement that built the primes, the government writes a specification, selects a contractor, and pays for labor and materials plus an agreed margin on top. The contractor's own capital is never at risk. Development risk sits with the taxpayer. Margins run in the range of 8-10%, and the only reliable way to grow the absolute number is to grow the cost it is calculated against. Anduril inverted every term. It funds its research and development with its own capital rather than the government's, more than 60% of revenue as of 2025, against low single digits at the legacy primes. It builds systems before a requirement exists, then brings a finished article to the customer for evaluation. And it sells at firm fixed price, which means the government knows what it is paying before work begins, and every dollar of overrun comes out of Anduril's pocket rather than the Treasury's. Gross margins run in the 40-45% range, considerably higher than the primes earn, which is exactly the point. The company makes more money by delivering faster and cheaper. That is the single most important incentive reform available in defense procurement. Private capital has validated the model on its own terms, moving Anduril from roughly a billion dollars in 2019 to $8.5 billion in 2022, $14 billion in 2024, $30.5 billion in mid-2025, and $61 billion this past May after a $5 billion round led by Thrive Capital and Andreessen Horowitz. Whether that number is correct is a question for investors. What it demonstrates for our purposes is simpler: the market is willing to fund American weapons manufacturing at enormous scale without the taxpayer underwriting the development risk, which the primes have insisted for decades was impossible.

Fury

The FQ-44A Fury is the clearest case of this. Anduril acquired the design when it bought Blue Force Technologies in September 2023 and redirected it from an air-combat training drone toward a full unmanned combat aircraft. The Air Force awarded a prototype contract in April 2024. Ground testing began in April 2025. First flight came on October 31, 2025, a clean-sheet design to a flight-tested aircraft in 365 days. On June 17, 2026, roughly four months ahead of schedule, the Air Force awarded production contracts for Collaborative Combat Aircraft Increment 1. Notably, it did not run the usual down-select that terminates the loser. It ordered both Anduril's FQ-44A Fury and General Atomics' FQ-42A Dark Merlin, more than 150 aircraft combined for delivery by the end of the decade, at a target unit cost under $30 million, roughly a third of an F-35A. Anduril's Ohio line is built to produce up to 150 aircraft a year in its present configuration, and the long-term service goal runs toward a thousand airframes. General David Allvin, the Air Force Chief of Staff, put the timeline plainly: these were aircraft that had been on paper less than two years earlier. That is the fastest path from prototype to production for a fighter-class aircraft in more than half a century. It was achieved by a company founded in 2017, competing against incumbents with sixty years of institutional advantage, and it was achieved because the incentives were pointed in the right direction.

The Motors

For this publication, however, the more significant work is happening in Stone County, Mississippi, and it has received a fraction of the attention. Solid rocket motors are the chokepoint of the American munitions base. Every interceptor, every guided rocket, every tactical missile requires one, and until very recently the entire United States had exactly two suppliers: Northrop Grumman and L3Harris. A two-supplier market for the single most essential input in the arsenal is not a market. It is a hostage situation, and it is precisely the vulnerability we identified in The Profiteer State when we argued that a nation which cannot produce the inputs to its own munitions has already surrendered a portion of its sovereignty. Anduril put $75 million of its own capital into the 450-acre Mississippi Solid Rocket Motor Complex at McHenry, built it out in eighteen months, and opened full-rate production in August 2025 — becoming the third domestic supplier of solid rocket motors in the United States. Annual capacity is moving from roughly 600 tactical motors to more than 6,000 by the end of this year. More than 700 motors have been test-fired since groundbreaking. The company has developed a proprietary aluminum-lithium propellant formulation, ALITEC, which it says extends motor range by up to 40 percent. The Defense Production Act Title III program has contributed $14.3 million in December 2024 and a further $43.7 million awarded in October 2025 — real public money, but structured as a minority investment alongside private capital that carried the larger share of the risk. That is exactly the arrangement this Program has called for: public capital used to close a strategic industrial gap, deployed as partnership rather than as a subsidy handed to a firm that would have collected it either way. Breaking a duopoly in energetics does more for American deterrence than a decade of white papers about the defense industrial base.

What Anduril Does Not Settle

Despite all of this, Anduril is still a defense contractor. It is valued at $61 billion. It maintains a registered lobbying operation and a corporate political action committee, and while its lobbying expenditures are a small fraction of what the Big Five spend, the direction of travel is the same one that produced the revolving door that the NNA condemned in July. It has essentially one customer (the US government, plus the allies Washington permits it to sell to) which is the structural condition that turns every large defense firm eventually into a political operation as much as a manufacturing one. Its valuation prices in near-flawless execution of a manufacturing bet that is, as of today, still young. Nor has everything it touches turned to gold. In April 2025 Anduril assumed the Army's troubled IVAS augmented-reality headset program from Microsoft by contract novation. This past July, a Congressional Research Service report confirmed that none of the 10,000 headsets developed and procured between 2018 and 2025 will ever be fielded, a $1.8 billion write-off. Most of that failure was inherited rather than created, as the Army's own assessment is that the system proved unaffordable to produce at scale. But a company that markets itself on succeeding where the primes fail should not get to quietly exempt the program where it did not. The remedies this Program has demanded do not become unnecessary because a contractor is currently performing well.

The Lesson

The reason this matters has nothing to do with any single company's fortunes. It matters because for forty years the American people have been told a story about why their defense industrial base decayed. The story held that this was simply what modern weapons cost, that the timelines were irreducible, that complexity demanded consolidation, that a serious nation could not build quickly anymore. It was a story told by firms whose revenue depended on it being believed, and it was believed, and on the strength of it a country that once built a Liberty ship in under five days came to accept twenty-three years as a reasonable interval between conceiving an aircraft and building it at rate. Anduril provides the refutation. Three hundred sixty-five days from a blank sheet to flying combat aircraft is the refutation. A third domestic rocket motor supplier standing up in 18 months on private capital is the refutation. All of this merely required a contract that rewarded delivery instead of delay, and a firm willing to put its own money at risk to prove that point.

Now, the Big Five have no excuse left. What remains is for Congress act. America does not have a manufacturing problem. It has a contracting problem. Anduril just proved it.