In June, Anduril CEO Brian Schimpf basically pulled the fire alarm on defense tech. Onstage at Fortune’s Brainstorm Tech conference, he said there “is a bit of a bubble,” explaining that when investors chase successful companies, “there can be very risky behavior,” and valuations can outpace what startups can realistically deliver.

That warning lands in the middle of a fundraising stampede. In the first quarter of 2026, VCs deployed a record $19.8 billion into defense tech across 262 deals, according to PitchBook. Back in Q1 2024, the figure was $5.7 billion, and in Q1 2025 it was about $17 billion. Defense has gone from Silicon Valley’s awkward side quest to a mainstream growth narrative, powered by AI-fueled innovation and the belief that this market is recession-resistant.

The twist is that “recession-resistant” does not mean “friction-free.” Defense tech has always been shaped by procurement reality, not just product-market fit. A defense company lives or dies by so-called programs of record, which are “notoriously difficult” to attain Pentagon budget line items. Even if a startup looks great in a demo, it still has to win through a bureaucratic contracting process and compete against defense “primes” like Northrop Grumman and Lockheed Martin, companies with entrenched positions and billions in contracts.

Over the last 20 years, the investor mindset shifted. The marked success of Palantir, eventually, and Anduril, more recently, showed venture-backed companies could emerge as serious players by going all-in on defense. Aaron Jacobson, a partner at VC firm NEA, put it bluntly: “Ten or fifteen years ago, it was really hard to get confidence that these companies would be able to get there.” Now, he argued, government progress and the visible wins make it easier to believe innovation in the space is feasible.

But feasibility is not the same thing as speed. Several sources in the piece stress that defense tech runs on “entirely different time scales and metrics” than traditional software. Morgan Hitzig of Overmatch Ventures described defense tech as almost the inverse of AI or SaaS economics. Software startups often start with low OpEx and then increase sales and marketing spend to scale. Defense tech, especially hardware-heavy efforts, requires significant CapEx upfront. Once a solution is deployed and proven, sales efficiency can become “extraordinary,” but getting to that stage can involve long delays.

That long runway is the heart of the “valley of death” problem. Janelle Teng Wade of Bessemer flagged that sales cycles are “long and lumpy” and hard to forecast, and that funding cycles are influenced by political factors and opaque contracting processes. Even for companies with military contracts, the status can remain pre-program-of-record, which keeps the future uncertain. Vantor’s Peter Wilczynski framed the big strategic question as funding across that valley: “I think the big question in defense tech is: Who’s going to fund it across that valley?” The worry for boards is simple: markets can re-rate quickly, but defense timelines do not.

So what happens when hot money meets hard procurement? Trae Stephens, Anduril cofounder and a Founders Fund partner, argues the supply of venture capital is now creating a distorted cycle. Funds compete to win deals, and their “best lever” is price, which leads to “chasing, chasing, and chasing.” Stephens emphasized the difference versus consumer AI: this “isn’t consumer AI,” and the growth curve is not “zero to $30 billion in 18 months.” Overvaluation can be survivable only for a while, because the business expectations eventually catch up. Stephens warned that investors might not care when there are no production contracts yet, but “at some point, that music’s going to stop” and the key question becomes who can avoid a down round when expectations are already priced in.

The overcrowding question is not abstract either. Ali Javaheri, senior emerging technology research analyst at PitchBook, says categories like battlefield AI and drones are looking especially saturated. He links the current bubble debate to attention on older models, implying the market can get fixated on hype patterns rather than procurement and deployment realities. And even if this does not mean a general implosion, it still points to a messy adolescent phase for the sector. Wilczynski suggested defense tech is likely “close to the adolescent period,” and that the next phase will be “awkward.”

For executives and investors, the strategic stakes are immediate: valuation is not just a number when financing bridges long procurement gaps. If capital stays cheap, teams can build faster and raise more, which may look like momentum. But if expectations outrun delivery timelines, boards can get trapped between underpowered commercialization and market re-rating risk. The core question peers need to answer is not whether defense tech is real. It is whether their sub-sector, capital stack, and milestones are resilient enough to survive the procurement timeline when the funding cycle eventually tightens.