War Is the Best VC Pitch Nobody Wants to Give

I didn't plan to write about defense tech. I build web apps. Laravel, Vue, the usual. The closest I get to military hardware is occasionally debugging a TYPO3 site for a client who makes industrial sensors.

But I've been staring at the funding numbers for the past two weeks, and something is happening in defense tech that I think the broader tech world is either ignoring or hasn't fully processed. So here's what I've pieced together, with the caveat that I'm a software engineer reading defense procurement contracts, not a policy analyst.

The Numbers That Made Me Open a Spreadsheet

On March 31, Saronic — an Austin-based company that builds autonomous warships — closed a $1.75 billion Series D at a $9.25 billion valuation. Led by Kleiner Perkins. Their previous round was $600 million at $4 billion. That's a 131% valuation jump in about a year. The company has 1,300 employees and a $392 million Navy production contract for something called the Corsair, a 24-foot autonomous vessel.

Five days earlier, on March 26, Shield AI raised $2 billion — $1.5 billion in Series G equity led by Advent International, plus $500 million in preferred equity from Blackstone. Valuation: $12.7 billion, up from $5.3 billion a year ago. Their core product is Hivemind, an AI pilot that lets drones fly autonomously without GPS. The V-BAT drone has flown over 130 combat sorties in Ukraine through Russian electronic warfare. They simultaneously acquired Aechelon, a simulation company.

And on March 14, the U.S. Army awarded Anduril a contract with a ceiling of $20 billion over ten years for their Lattice AI platform, focused on counter-drone capabilities. I want to be precise here: $20 billion is the ceiling, not a guaranteed spend. The first actual task order was $87 million, issued March 16. But even so — $20 billion ceiling, single vendor, for a company that didn't exist eight years ago.

That's $23.75 billion in combined deal value across three companies in eighteen days.

I opened a spreadsheet. Then I kept going.

How We Got Here

According to Dealroom, global defense-tech startup funding was $869 million in 2020. In 2025, it hit $11.2 billion. That's more than a tenfold increase in five years. CNBC cited these exact numbers in their April 3 newsletter. Other tracking platforms report higher totals — PitchBook has $49.1 billion for 2025 — but that's because they count dual-use companies more broadly. The Dealroom figures are the narrower, defense-specific slice.

Either way, the direction is unmistakable.

Some of this was already in motion before the Iran war. Anduril raised $2.5 billion at a $30.5 billion valuation in June 2025, led by Founders Fund — their largest single check ever, at $1 billion. Helsing, the Munich-based defense AI company chaired by Spotify's Daniel Ek, raised €600 million at a €12 billion valuation the same month. Ten new defense-tech unicorns were created in 2025 alone.

But then Operation Epic Fury started on February 28, 2026, and the funding environment went from warm to molten.

What the War Changed

I need to talk about Palantir's Maven Smart System, because it's central to understanding why VCs are piling into this space.

Maven — originally Project Maven, the Pentagon's AI initiative that famously caused Google employees to revolt in 2018 — is now a Palantir-run targeting platform that consolidates over 150 data sources into a single operational picture. Satellites, drone feeds, intercepted communications, all of it. On day one of the Iran campaign, Maven presented over 1,000 strike options. CENTCOM Commander Admiral Brad Cooper said the technology "effectively doubled both the pace and intensity of military strikes." The Pentagon's Chief Digital and AI Officer, Cameron Stanley, said at Palantir's AIPCon 9 on March 12: they've gone from identifying a target to actioning it, all from one system.

The U.S. struck over 11,000 targets total. Palantir's stock is up accordingly. CEO Alex Karp went on CNBC to say the platform gives the West "a critical edge." He declined to comment on whether Maven was used to target Khamenei specifically.

Here's why this matters for the funding landscape: Maven proved, in a live combat environment watched by the entire world, that AI-driven targeting infrastructure works. Not in theory. Not in a demo. In an actual war. Every VC who was on the fence about defense tech watched that happen and reached for their checkbook.

The Pentagon is now spending $13.4 billion specifically on AI and autonomy in FY2026 — the first time there's been a standalone budget line for it. And on April 3, Trump submitted a budget request of $1.5 trillion for defense — $1.15 trillion base plus $350 billion from a reconciliation bill that hasn't passed yet. I should note: this is a proposal, not law. Congress has to approve it. But the signal is clear.

The Part Nobody Wants to Say

Here's where I start feeling uncomfortable, and I think that discomfort is worth noting rather than papering over.

Defense-tech startups are building impressive technology. Saronic's autonomous ships are real. Shield AI's Hivemind is flying real combat missions. Anduril's Lattice is processing real intelligence data. These are not vaporware companies with pitch decks and no product. The engineering is genuinely good.

And also: the reason the money is flowing is because people are dying.

CNBC ran a newsletter on April 3 titled "Defense startups eye Iran war windfall." I don't think "windfall" is the wrong word. On March 31, the same day Saronic closed its $1.75 billion round, Iran's Revolutionary Guard Corps named 18 U.S. tech companies as "legitimate targets" — Apple, Google, Microsoft, Nvidia, Meta, Tesla, Palantir, and eleven others. The IRGC explicitly called out AI companies as "the main element in designing and tracking assassination targets."

There are over 3,400 Iranian casualties. Thirteen U.S. service members have been killed. The Pentagon is investigating Maven's potential role in a strike on an Iranian girls' school that killed roughly 170 people. The Strait of Hormuz is closed. Oil is over $110 a barrel.

And the funding keeps accelerating.

I'm not making a moral argument here — or maybe I am, I'm not sure. I'm an engineer. I notice when incentive structures point in a direction that should probably give people pause, and nobody seems to be pausing.

The European Angle

Since I'm based in Europe, I'll add this: the defense-tech boom isn't just American.

Germany's defense budget hit roughly €108 billion in 2026 — a €20 billion increase from 2025. Chancellor Merz has pledged €650 billion over five years to make the Bundeswehr the strongest conventional army in Europe. The EU's ReArm Europe plan aims to mobilize €800 billion in additional defense spending by 2030. European defense-tech VC went from €200 million in 2021 to €2.6 billion in 2025, according to McKinsey.

Helsing, out of Munich, is now valued at €12 billion. European defense startups raised $854 million in Q1 2026 alone. CNBC reported on March 31 that European defense startups are actively hiring in the Gulf and pitching Middle East governments on commercial deals tied to the Iran conflict.

The money is flowing on both sides of the Atlantic. The question of who it's flowing toward, and what those companies are building, and who decides how it gets used — those questions are moving a lot slower than the capital.

What I Keep Coming Back To

I started this piece because the numbers were striking. $1.75 billion for autonomous warships. $2 billion for autonomous drones. A $20 billion contract ceiling for counter-drone AI. $13.4 billion in Pentagon AI spending. A $1.5 trillion defense budget proposal. Defense-tech funding up twelvefold in five years.

But the number that actually stuck with me is different. In 2018, Google employees signed a petition — about 4,000 of them — demanding the company withdraw from Project Maven. Google did. It was seen as a turning point: tech workers drawing a line on military AI.

Maven is now the backbone of the largest U.S. military campaign since Iraq. It has presented over 1,000 targeting options in a single day. It is run by Palantir, backed by every major VC firm, and staffed by engineers who presumably applied voluntarily.

Eight years. That's how long it took for "we won't build this" to become "this is revolutionary."

I'm not sure what to do with that. But I think it's worth saying plainly.