Defense Unicorns: How Ukraine, Drones, and $49 Billion in VC Minted 2025's Most Unusual Startup Class

A single year produced more billion-dollar defense startups than the previous decade combined, as a generation of Silicon Valley founders decided the most urgent problem to solve was war.

By Carry and Conquer Publications

The numbers from 2025 are almost disorienting. Venture capital poured more than $49 billion into defense-related startups over the course of the year, according to PitchBook - a figure that rivals only frontier AI in terms of sector-wide momentum. Ten new defense tech unicorns emerged in twelve months. The sector's combined active unicorns are now collectively valued at roughly $495 billion, more than double their 2021 worth. This is not a trend. It is a structural realignment of one of the world's largest and least disrupted industries, and it was largely triggered by 117 cheap drones launched from cargo trucks deep inside Russia.

What Ukraine Changed

On June 1, 2025, Ukraine's Security Service executed a covert drone operation that military analysts are still studying. Codenamed Operation Spiderweb, it had been in planning for over a year and a half. Operatives had smuggled 117 first-person-view drones into Russia, hidden inside the wooden-roof cabins of cargo trucks, and parked them near five air bases spread across five time zones - including one base, Belaya, located more than 4,300 kilometers from the Ukrainian border in Eastern Siberia. At a designated moment, the rooftops opened remotely. The drones flew out and hit their targets simultaneously. According to Ukrainian sources, 41 aircraft were destroyed or disabled, including Tu-95 and Tu-22M strategic bombers. The estimated damage: $7 billion. The estimated cost per drone: a few thousand dollars.

The operation was immediately compared to Israel's 2024 pager attacks on Hezbollah - a precision intelligence-driven strike designed to cause maximum material damage at asymmetric cost. But the implications for the defense industry went beyond the battlefield optics. For the investor community, Spiderweb was a proof point for the thesis that had already been building for two years: cheap, software-guided, mass-produced autonomous systems can defeat vastly more expensive legacy hardware. Every dollar of damage in that attack was inflicted using commercial off-the-shelf components, open-source navigation software, and AI trained on museum aircraft. The defense industry's century-old assumption - that better weapons meant more expensive weapons - had just been publicly invalidated at scale.

The investors who had already placed their bets responded accordingly. The ones who had not began writing checks.

Ten New Unicorns

The Forbes tally of 2025's defense unicorn class shows how different the new wave looks from the companies that preceded it. Where Anduril, Palantir, and SpaceX built early moats over years of quiet groundwork, 2025's new entrants moved faster, across more categories, and with more aggressive valuations than anything the sector had seen.

Chaos Industries - founded in 2022 by John Tenet and Bo Marr, who had previously built counter-drone unicorn Epirus - raised $510 million in a Series D in November at a valuation of $4.5 billion. The company builds coherent distributed radar networks capable of detecting airborne threats including drones, missiles, and aircraft up to ten minutes faster than legacy systems. Its Hawthorne, California campus became one of the year's most-watched defense addresses.

Castelion, founded by former SpaceX executives Bryon Hargis, Sean Pitt, and Andrew Kreitz, raised $350 million in Series B funding in December, led by Altimeter Capital and Lightspeed Venture Partners. The El Segundo, California company's pitch is simple: the United States is losing the hypersonic weapons race to China, and the only way to catch up is to apply the SpaceX manufacturing playbook - build fast, test constantly, drive down unit costs - to tactical missiles. In 2025, Castelion conducted more than 20 development flight tests validating subsystems for its first weapon, the Blackbeard missile. The company raised funds to build Project Ranger, a New Mexico facility designed for high-rate production of thousands of rounds annually. The U.S. Army included Castelion in its FY2026 budget request with a $25 million line item for Blackbeard development.

On the European side, Quantum Systems - a German-Ukrainian drone company founded in 2014 by Florian Seibel, a former German army pilot - raised a EUR 160 million Series C led by Balderton Capital, crossing the billion-dollar threshold in May. The company's Vector reconnaissance drone had already accumulated years of combat use in Ukraine before investors took notice; after Russia's 2022 invasion, Quantum Systems became one of the first drone manufacturers to establish full production operations inside Ukraine itself. Their plant in Ukraine can now produce one Vector drone per day and has repaired roughly 200 systems and trained 350 pilots.

Portugal's Tekever followed the same combat-validation path to unicorn status. Its AR5 long-endurance surveillance drone has flown over 10,000 combat hours in Ukraine, according to the company, and has contributed to an estimated $4 billion in damage to Russian military assets, including two S-400 air defense systems. The NATO Innovation Fund, Baillie Gifford, and Ventura Capital backed Tekever's round, which pushed the company's valuation past $1.25 billion.

Govini, a Pittsburgh-based AI analytics company led by CEO Tara Murphy Dougherty, reached unicorn status in October after Bain Capital invested $150 million. The company's Ark platform tracks and analyzes defense supply chains down to raw materials - flagging, for example, when a sub-tier parts supplier shows signs of bankruptcy risk before it disrupts a program. Govini reported $100 million in revenue for fiscal 2025 and a 300-person workforce. It is already embedded in the U.S. Army's 4th Infantry Division, where it tracks ammunition consumption in real time and generates AI-based resupply forecasts.

Forterra, Destinus, Impulse Space, Apex, and Uforce - a London-based startup built from nine Ukrainian defense companies and valued at $1 billion after raising $50 million in seed funding in early 2026 - round out the picture of a class that spans continents, categories, and timelines.

The Investment Logic

For decades, venture capital avoided defense. The sector was considered slow, ethically fraught, dependent on multi-year procurement timelines, and ultimately a poor fit for VC's exit-oriented business model. The attitude shift was gradual, then sudden. Russia's invasion of Ukraine in 2022 began moving the conversation. The real pivot came when the category's existing winners - Anduril, Palantir, SpaceX - started posting results that made the sector credible to generalist investors.

By 2025, the transformation was complete. Point72 launched dedicated defense funds. Y Combinator began actively seeking defense startups. Andreessen Horowitz had already built its American Dynamism practice around national security. Nearly one in twelve VC dollars globally was flowing into defense tech by the end of the year.

PitchBook data shows the median VC defense tech valuation in 2025 hit $146 million, compared to $42.8 million the prior year and $22.8 million across all sectors. Nearly 60% of all defense tech funding went to drone startups. Mega rounds drove the headline numbers: Anduril raised $2.5 billion in June at a $30.5 billion valuation, with Founders Fund writing the firm's largest single check - $1 billion - to lead the round. Helsing, the German AI defense company, raised EUR 600 million in the largest single round in European defense tech history. Saronic, which builds autonomous surface vessels for the Navy, raised $600 million. Defense spending across NATO nations rose 9% in 2025, passing $2.7 trillion globally in the sharpest annual increase in over three decades.

Secretary of Defense Pete Hegseth's "speed to capability" mandate gave the investment case a political tailwind. For the first time, the Pentagon was openly signaling that it would prioritize fast procurement from innovative startups over legacy contractors with overrun-prone development timelines.

The Case Against the Boom

Not everyone reads the moment as clearly as the capital flows suggest.

Connor Love, a partner at Lightspeed Venture Partners and investor in the sector, acknowledged the risks directly: "We can argue about a bubble - there's going to be a lot of people putting money in the wrong things and a lot of overvalued companies." PitchBook analysts identified drone startups in particular as showing signs of overheating, with nearly 60% of defense funding concentrated in a category where five of the nine existing unicorns operate, where many companies lack differentiation, and where established players like Anduril and Shield AI already hold strong contract positions.

The critical structural problem is the gap between valuation and revenue. Most 2025 defense unicorns have not yet won the large, multi-year procurement contracts that would justify their prices. The model - raise capital at a tech-company multiple, then wait years for the government to place orders - depends on the DoD actually following through on its reform promises. History offers little comfort: the "valley of death" between prototype and production has killed off generations of well-funded defense tech startups. A December 2025 analysis by Defense Tech Signals put the challenge plainly: "What remains is harder and far less forgiving: turning awarded demand into reliable, repeatable capability."

Critics from outside the investment community raise a separate set of concerns. A report from Responsible Statecraft argued that VC's structural incentives - build fast, raise valuations, exit before the long tail of military procurement plays out - are fundamentally misaligned with the defense market's actual dynamics. Venture capital is built for mobile assets that reach exit on a startup timeline. Pentagon contracts are built for permanent infrastructure, decade-long supply chains, and suppliers that stay in the business. The capital is entering the sector because it looks like AI. The returns will be determined by whether it can function like a defense contractor.

What Comes Next

The signal investors are watching is contract conversion. Over the next 12 to 18 months, several of 2025's unicorns are likely to begin winning or losing the large national contracts that will separate the lasting players from the overvalued bets. Castelion's Blackbeard missile needs to pass Army platform integration testing before a full production contract can be awarded. Quantum Systems' Vector drone needs to scale European production beyond its current single-aircraft-per-day capacity in Ukraine. Govini's Ark platform needs to demonstrate that its supply chain analytics can sustain and expand its Army contracts into a full enterprise deployment.

Three companies graduated to the Defense News Top 100 list in 2025 - SpaceX at number 40, Palantir at 70, and Anduril at 93. These are the proof of concept that the rest of the class is trying to replicate: defense tech startups that have successfully crossed from VC-backed experiment to recognized defense contractor. The distance between where most of 2025's unicorns stand and where those three are is still enormous.

But the structural conditions that produced the 2025 class are not going away. Ukraine has become a continuous real-world testing environment for every drone, autonomy, and sensing technology on the market. Chinese capabilities in hypersonics and autonomous systems have created a genuine national security gap that legacy contractors cannot close on their current timelines. Global defense budgets are rising across NATO, driven by a combination of genuine threat perception and the political reality of reduced U.S. security guarantees in Europe.

The money bet in 2025 is that a small number of the new class will follow Anduril and Palantir's path - from technically interesting startup to irreplaceable supplier. Lightspeed's Connor Love put the expectation at five or six lasting winners who change the entire landscape. Whether the other four or five produce sustainable businesses, or whether they produce the writedowns that historically follow any boom cycle in capital-intensive hardware markets, will be determined in the contracting offices of the Pentagon over the next several years. The battlefield has already voted.