Anduril and Mach are soaring, but Ross Fubini sees a valley of death

Defense tech is having a moment that would have sounded absurd not long ago. Anduril and Mach Industries have just doubled and quadrupled their valuations, respectively, while the U.S. government is proposing a 40% increase in defense budget. That is the kind of capital and policy backdrop that makes founders, investors, and operators lean in. It also explains why a fresh crop of startups is racing into the sector, hoping to turn prototype work into real federal business before the window closes.

But Ross Fubini, the venture investor who wrote Anduril’s first check, is not buying the easy version of the story. His view, as reported here, is blunt: most of these companies will get lost in the Valley of Death between prototype contract and something more durable. In other words, getting a first shot from the government is one thing. Turning that into a scaled, lasting business is another. For anyone building in defense tech, that gap is the whole game.

That gap matters because defense procurement is not the same as selling software to a startup or cloud service to a Fortune 500 company. Government contracts tend to move through layers of validation, compliance, and long sales cycles, and the path from a promising demo to repeatable revenue is usually far less forgiving than in private markets. The source does not spell out the mechanics in detail, but Fubini’s warning points straight at the industry’s central challenge: there can be a big difference between a company that can impress in a prototype and one that can survive the grind of becoming a supplier the government can depend on.

The money flooding into the category makes that challenge easier to ignore, which is exactly why it matters. When valuations are climbing and the budget outlook looks friendlier, startups can look as if they are all riding the same wave. But waves are not business models. A 40% proposed increase in defense budget signals bigger opportunity, yes, but it also signals a more crowded field of companies all chasing the same contracts. In that environment, the winners are unlikely to be the loudest or the fastest to raise. They are more likely to be the ones that can navigate procurement, withstand the delays, and keep building after the initial excitement fades.

That is what makes Fubini’s perspective especially useful for executives and boards watching the sector. He is not speaking as an outsider skimming a hot trend. He is the investor who wrote Anduril’s first check, which gives his warning some real texture. If one of the category’s breakout names can now double in valuation, that does not mean the entire landscape is healthy in the same way. It means the market is starting to separate companies with real staying power from companies benefiting from momentum, narrative, and a funding stampede.

For founders, the message is pretty clear even if the source is short on specifics: defense tech is drawing attention, but attention is not the hard part. The hard part is becoming built to last in a business where the customer is the government and the timeline is never entirely in your control. For investors, that means underwriting more than growth dreams. It means asking whether a startup can move beyond the prototype phase and into the deeper, slower work of actual deployment and contract durability. For established players, it means the influx of capital may increase competition faster than it improves the odds for everyone.

The bigger implication is that defense tech may be entering a classic sorting phase. Surging valuations and a larger budget create the appearance of broad opportunity, but Fubini’s warning suggests the category will not reward everyone equally. The startups most likely to matter are the ones that can cross the Valley of Death, not just get to the edge of it. That is the strategic line to watch now: not who can raise in a hot market, but who can turn a prototype into a platform, and a platform into a business that lasts.