Consider 2018: Googlers were storming out over their company’s involvement in the AI military initiative Project Maven, and Anduril was an anomaly, a multimillion-dollar defense-focused startup soon to be headlined as “the most controversial startup” in tech. VCs touched defense rarely, if at all.

Today, they can’t get enough. Anduril, now valued at $61 billion, is joined by a growing class of “neo-primes,” from autonomous shipbuilder Saronic, last valued at $9.25 billion, to drone maker Shield AI, at $12.7 billion. Defense has exploded into a consensus growth area among VCs, seen as ripe for AI-fueled innovation. The numbers bear it out: In the first quarter of 2026, VCs deployed a record $19.8 billion into defense tech across 262 deals, according to PitchBook. (For comparison: That number in Q1 2024 was $5.7 billion, and in Q1 2025 was about $17 billion.)

The vibe shift is sending valuations into the stratosphere. Early-stage defense startups are raising millions and fetching multiples from 17 times to 50 times revenue (sometimes even higher). 

With the market running so hot, the inevitable question that emerges: Are we in a bubble? At the Fortune Brainstorm Tech conference in June, Anduril CEO Brian Schimpf answered with a nuanced but decisive “Yes.” 

“When there are successful companies, you have lots of other companies and investors chasing that, and [there can be] very risky behavior,” Schimpf said onstage. “We’ve been very careful at every stage to manage this, but it’s easy to chase those valuations if you’re not being careful. So, I do think there’s a bit of a bubble. Capital’s cheap, and it’s great if you can get capital as an advantage—go for it. But do understand the consequences for your company when you do: You put expectations through the roof of what you can deliver on.”

Given how far ahead valuations have gotten from the fundamentals of many startups, it’s tough to argue that there’s not a bubble in at least some corners of the defense tech market. But that’s not to say these aren’t real businesses or that the market is on the verge of imploding.   

“Is there a bubble in defense tech? It’s always a hard question,” said Peter Wilczynski, chief product officer of spatial-intelligence software maker Vantor. “It’s obvious to say yes. But the people who say yes then are always wrong, so it’s complicated.”

I think the more interesting series of questions goes something like this: What parts of defense tech funding, as it exists today, are fragile, and what is likely to prove resilient? Are traditional Silicon Valley investors actually equipped, in mindset and business models, to underwrite and ride out the gulf between defense (and D.C.) economics versus the economics of consumer apps or enterprise software? And, if there is a bubble, is that actually bad? 

No matter what, there’s certainly friction ahead. 

“We’re probably close to the adolescent period, for sure,” said Wilczynski, who previously spent 12 years at Palantir. “There are going to be teenage years in defense tech, and it’s going to be awkward.”

The unforgiving economics of defense

For decades, investing in defense companies and weapons makers wasn’t just déclassé: There was limited evidence it could work. A defense company lives or dies by so-called programs of record—a notoriously difficult to attain Pentagon budget line item, bureaucratic but business-making or -breaking—and was up against the defense “primes” like Northrop Grumman or Lockheed Martin, known for their billions in contracts and market cap, their history, and entrenchment within Congress. 

But over the past 20 years, the perception among investors has radically shifted: The marked success of Palantir—eventually followed by Anduril—made clear that serious venture-backed winners could emerge by going all in on defense. 

“Ten or 15 years ago, it was really hard to get confidence that these companies would be able to get there,” said Aaron Jacobson, a partner at VC firm NEA. “But because you’ve now had success in Palantir and Anduril and others, plus the government’s progress, it’s starting to really encourage folks to believe it’s actually going to be possible to have more innovation in the space.” 

For companies that are able to get traction, the economics are alluring: large total addressable markets; the Pentagon as a long-term customer after committing to a purchase; and businesses that are seemingly recession-proof. That said, it also gets tough fast, Janelle Teng Wade, Bessemer partner, pointed out via email: Sales cycles are “long and lumpy” and “hard to forecast,” while you’re also stuck with “funding cycles impacted by political factors, and opaque contracting processes.”

The timelines are just different, and finding analogous industries to benchmark defense companies against is not easy.  

“Defense tech runs on entirely different time scales and metrics than traditional software,” said Morgan Hitzig, Overmatch Ventures general partner, via email. In fact, Hitzig explains, defense tech is in many ways the inverse of AI or SaaS economics. 

“Software companies start with low OpEx [operating expenditure] and then accrue high sales and marketing spend to scale,” Hitzig says. “Defense tech requires significant capex [capital expenditure] upfront, especially for companies building physical hardware. But once deployed and proven, sales efficiency can be extraordinary.”

VCs like to talk about patient capital, but defense tech requires really patient capital. The “valley of death”—the famously lengthy wait between tech that works and the military buying and deploying it at scale—is easily enough to kill a company. Many multibillion-dollar names in defense tech, like Shield, have military contracts, but remain pre-program-of-record, which keeps the future in limbo.

“The ‘valley of death’ is also about: How do you go from a cool startup to a full-fledged neo-prime that’s able to return cash to investors?” said Vantor’s Wilczynski. “I think the big question in defense tech is: Who’s going to fund it across that valley?”

The answer, for now, seems to be VCs. 

‘Too much supply’

Defense tech’s rise has been underpinned by surging geopolitical strife, from the war in Ukraine to the war in Iran. The question is whether there are too many companies raising at unsustainable valuations, and if venture capital is the right tool for tech that’s inextricably tied to national security.

“The challenge is that there’s too much supply in venture capital,” said Trae Stephens, Anduril cofounder and Founders Fund partner, who argues that the glut of capital can create a funding cycle that distorts the market. “All of these funds are competing with one another to win deals, and the best lever they have to win deals is price. They keep chasing, chasing, and chasing.”

The problem with that dynamic in defense tech, Stephens said, is that “this isn’t consumer AI. The growth curve isn’t zero to $30 billion in 18 months. It’s a really, really hard game.”

The long timelines make the high out-of-the-gate valuations tricky over time. 

“You could find yourself in the situation where you’re burning millions of dollars per month, and you still don’t have any production contracts,” said Stephens. “There’s enough hype right now that investors don’t really seem to care about that. But at some point, that music’s going to stop, and there are going to be real business expectations. The question is: In that game of musical chairs, when the music stops playing, who’s going to be in a good enough position in relation to their valuation that they can avoid a down round?”

Which sub-areas in defense tech seem overcrowded right now? Ali Javaheri, senior emerging-technology research analyst at PitchBook, says that categories like battlefield AI and drones are looking especially saturated. 

“The reason that people might kind of be fixated on this question—whether it’s a bubble or not—is because they’re still conceiving of the old model of buying defense tech,” said Javaheri, referencing the long-standing criticism that the Pentagon is a bureaucratic and snail-paced buyer of tech, and Secretary of War Pete Hegseth’s current efforts to streamline the procurement process. “This new model … It hasn’t shaken out yet. I think it’s still way too early to say whether it’s a bubble or not.” 

So, asked in good faith: Is there a bubble? Yes, probably, maybe, or eventually. But what’s more important is this: There will be ways in which a bubble isn’t a bad thing for defense—the hype-based capital fueling innovation right now could productively shape national security and financing norms for decades to come. The flip side—and a potential cause for concern—is that this is bigger than startup valuations. A period of deep disillusionment on the other side of a bubble burst could potentially affect the health of the U.S. defense innovation base.

For now, there’s lots of room to run in defense tech, for founders and investors. But, in the end, there will only be so many major winners. For investors, Stephens cautions: Picking the right horse is ultimately more important than timing the bubble. 

“If you’re a space tech investor and you didn’t invest in SpaceX, you probably lost money,” he told Fortune. “If you’re a crypto infrastructure investor and you didn’t invest in Coinbase, you probably lost money. If you’re a social media investor and you didn’t invest in Facebook, you probably lost money. We have this crazy tendency to believe that the emergence of a sector will lead to dozens of winners—and there’s just no evidence that’s actually how any of this works.”