The U.S. is waging a war with Iran and bringing to bear few weapons systems less than fifteen years old. The new systems include attack drones reverse-engineered from the Iranian Shahed and drone boats from startup Saronic. Most everything else is decades old, enormously expensive, and getting out-fought by a smaller arsenal that costs a fraction to build.
That is the war Defense Secretary Pete Hegseth’s acquisition revolution is producing. He is feeding a venture-capital bubble with defense dollars and calling it disruption.
The numbers tell the story the talking points won’t. Pentagon contract spending on the fifteen highest-valued defense-tech startups tripled between 2022 and the last fiscal year. Sounds like revolution. Then you read the next line: those startups still account for less than one percent of total defense contractor dollars. The primes — Lockheed, RTX, Northrop Grumman, General Dynamics, Boeing — doubled their take to $372 billion over the same period. The startups got $4 billion out of $506 billion total. Hegseth doubled the old guard while pitching himself as the man who would break their monopoly.
Eisenhower, in his farewell address of January 17, 1961, warned of “the acquisition of unwarranted influence, whether sought or unsought, by the military-industrial complex.” Sixty-five years later, the complex has not shrunk. It has metastasized into a two-tier system where the primes eat the big dollars and a swarm of venture-backed startups compete for scraps, while the investors who back them harvest returns on Pentagon contracts the way they would any other speculative asset.
The investor exuberance is not subtle. Anduril Industries doubled its valuation from $30.5 billion to $61 billion in a year. Venture-capital investment in defense and aerospace startups hit $16.8 billion in the first half of 2026 alone, exceeding any prior full-year total. More than 400 drone companies now operate in the United States. Roughly 10,000 new defense companies have entered the market in two years. Trae Stephens, Anduril’s co-founder and a partner at Founders Fund, said on a recent podcast that he is “very uncomfortable” with the moment, that “prices are untethered from reality.” When the man building the weapons says the price tag is fantasy, that is not a contrarian take. That is a warning from the factory floor.
What Hegseth has actually done is strip the guardrails. He eliminated the requirements process that had governed weapons purchases — slow and cumbersome, yes, but also the mechanism that was supposed to ensure a weapon worked before the Pentagon bought it. He suspended costly cybersecurity requirements. He pushed buying authority down to lower-ranked officers. He stood up programs catering to startups and sped up contracting. “We’ve been waging a war of attrition against the Pentagon bureaucracy,” he told the Journal. A war of attrition against your own acquisition guardrails is not reform. It is clearance for whoever gets there first.
And who gets there first? Not the small startup in a garage. Anduril and Saronic — two of the deepest-pocketed defense-tech companies — accounted for about a quarter of all Pentagon contract spending to the startup cohort last year. The mechanism is not mysterious. The same venture-capital networks that funded Anduril and Saronic installed them on the Pentagon’s preferred-vendor lists; the same political connections that opened doors at the Defense Innovation Unit now open doors at the new autonomous warfare unit. The rest of the startup cohort — the hundreds of firms the rhetoric promised to lift — splits the remainder and watches lawsuits pile up as smaller companies accuse the military of playing favorites. The disruption Hegseth promises looks, on closer inspection, like a venture-capital oligopoly wearing a hardhat. The names have changed. The procurement channel is the same.
The structural problem this disguises is older than Hegseth. The 1993 “Last Supper” meeting, where Deputy Secretary of Defense William Perry told defense executives to consolidate or starve, reduced roughly fifty primes to five. Bacevich, in Washington Rules, named the result: a bipartisan consensus that American military preeminence requires permanent global deployment backed by permanent arms spending. Hegseth has not challenged that consensus. He has expanded its beneficiary list — adding venture capital to the check that used to go only to the primes — while keeping the check itself at historically unprecedented size.
The $1.5 trillion budget request makes the picture plain. Congress is resisting it, and not only because of the number. Officials from both parties have demanded scrutiny of Pentagon spending on loans and equity stakes in startups, many backed by venture capitalists including Donald Trump Jr. When the defense secretary’s acquisition shake-up overlaps with a presidential family member’s investment portfolio, the conflict-of-interest that Eisenhower warned about does not always arrive with a lobbying registration. Sometimes it arrives with a fundraising round.
The Iran war made the gap visible in real time, and Hegseth’s bubble responded the way bubbles do. The few new systems — Shahed-derived drones, Saronic drone boats — got the press. Most everything else the U.S. has brought to bear is decades old and enormously expensive. Iran’s smaller and lower-cost arsenal has denied the U.S. military total victory. A war that was supposed to showcase the new defense-tech ecosystem instead showed that the ecosystem has not yet produced the systems that fight the war.
Michael Brown, an early leader of the Pentagon’s Defense Innovation Unit, gave the department an “incomplete” on transformation. The real test, he said, is whether Congress passes budgets that benefit startups. But that framing concedes the point: if the reform depends on Congress allocating billions to unproven companies while the primes still eat the majority, then the reform is a budget line, not a structural change. The acquisition bureaucracy that has resisted reform since 1960 has not been dismantled. It has been routed around, temporarily, by a secretary who suspended the rules and called the suspension progress.
What the country actually needs is an off-the-shelf acquisition pipeline — one that can buy proven commercial technology, adapt it for combat, and field it in months rather than the decade-plus a traditional program of record requires — without surrendering the safety and cybersecurity reviews that keep broken weapons out of service members’ hands. That is a harder project than cutting checks to Silicon Valley. It requires institutional reform, not institutional suspension. And it requires an honest accounting of where the money goes — which is still, overwhelmingly, to the same five companies that have been eating the defense budget since the Last Supper, amid a federal deficit approaching $1.9 trillion.
Eisenhower’s warning was not about who gets the contract. It was about the concentration of power that results when a permanent war establishment becomes a permanent economic constituency. A venture-capital layer on top of the primes does not disperse that power. It doubles the number of voices in the room demanding more money, while the men and women in uniform — and the citizens who pay for all of it — get a system that still cannot field new weapons faster than the adversary can build cheap ones.