TheChek News Flow

Defense tech companies argue that warfare is changing rapidly—but the current conflict is using mostly traditional weapons.

War focuses attention on the present. That was very clear from a week’s worth of conversations in Washington with leaders at defense tech startups, traditional military contractors and people close to the Pentagon.

Until the U.S. and Israel attacked Iran, buzz was building around startups such as Anduril and Saronic that promised to deliver AI-infused autonomous weapons quickly and economically. The U.S., meanwhile, has relied more in the current conflict on cruise missiles and bombs from its regular stable of traditional defense contractors.

“It’s not very different from how we would have fought a war a decade ago, two decades ago, in terms of munitions,” said Roger Zakheim, director of the national security–focused Ronald Reagan Presidential Foundation and Institute.

That’s a reality check for the venture capital firms that have invested billions of dollars under the premise that warfare is changing rapidly and new defense tech can disrupt old-school contractors. But wartime typically requires only the best-tested weapons systems, and many of Anduril’s most notable offerings, like its autonomous fighter jet, are still just prototypes.

Instead, today’s war has illustrated the importance of prime contractors like RTX, which owns Raytheon, maker of Tomahawk cruise missiles, and Lockheed Martin, builder of fighter jets. Their equipment is in high demand, and restocking supplies will give them business for years.

Moved to the background are startups raising a growing pile of VC dollars to try to win new military programs, a total that reached $56 billion last year, up 83%, according to the Reagan Institute. But that number far exceeds the amount of money private defense tech companies earned in Pentagon contracts: $4.3 billion last year, less than 1% of that total.

The startups are trying to follow in the footsteps of companies such as SpaceX and Palantir, whose technology already permeates the Pentagon. Aside from Anduril, which aims to make autonomous fighter jets and other weapons, some of the best funded are Saronic Technologies, building autonomous naval boats; Shield AI, developing AI piloting software for autonomous aircraft; Neros, which makes drones; and Castelion, which is building hypersonic weapons.

These companies are trying to wedge their way into a defense budget that’s largely committed to existing programs, said Tara Murphy Dougherty, who runs Govini, a Bain Capital–backed software company that helps the Pentagon track its weapons arsenal. “We’re not gonna stop funding the fifth-generation fighter jet or nuclear submarines. You’re left with a much smaller [budget for startups] than people realize,” she said.

Some startups are aiming to fill niches made obvious by the war in Ukraine, including cheap drones that traditional contractors can’t build.

One company getting increased interest is Heven AeroTech, a Virginia-based startup making hydrogen-powered drones, valued at over $1 billion. Bentzion Levinson, a former Israeli military commander who is the startup’s CEO, said the military’s needs have changed. “Drones that can drop things or launch things or kamikaze drones are really the future of this ecosystem,” he said.

But it’s still a slog for startups to win the government’s trust. Levinson’s company announced a “basic ordering agreement” with the U.S. Army for its drone last week, a type of contract that doesn’t require the military to actually buy the drones but speeds up the process if it decides to.

The Iran war, meanwhile, has given the U.S. a glimpse of how adversaries are using cheap, autonomous weapons in battle. That includes Iran’s Shahed Aviation Industries, whose low-cost drones have become its crucial weapon. “It’s far more the adversaries’ Anduril than our Anduril,” said Zakheim, referring to Shahed.

I spent last Thursday at a national security conference held by the Reagan Institute, a more establishment affair sandwiched between the VC-oriented defense tech conferences put on this month by leaders at Andreessen Horowitz and Founders Fund. Instead of Palantir’s fiery CEO, Alex Karp, who headlined Andreessen’s American Dynamism Summit, the Reagan Institute featured Michael Duffey, the low-key undersecretary of war for acquisition and sustainment.

Duffey called for more private investment in defense, but said the money should go to build more factories, finance component makers and fund supply chain resilience, even if “it’s not as headline grabbing as a new prototype and technology.” The institute gave out grades for how the U.S. was meeting its national security goals. (The quality of new defense tech startups got one of the highest grades, A–, but overall “defense modernization” received a D and manufacturing capacity received D+.)

The good news for at least some prominent newer defense firms and their investors is that military leaders want to help defense tech startups succeed. “How do we get barriers out of the way to find pathways forward for the Saronics, Anduril, Shield AI to get access to those opportunities?” Duffey said.

Anduril, for instance, won what could become its largest contract to date last week. The deal to provide software and other products to the U.S. Army could reach up to $20 billion over 10 years, although the money isn’t fully committed. Getting more such deals would support Anduril’s projections of hitting $16 billion of annual revenue by 2030, up from $2.2 billion last year.

But the vast majority of defense startups won’t “punch all the way through” to become large sellers of technology to the Pentagon, acknowledged Alex Moore, a defense tech investor at 8VC and a board member at Palantir. He said traditional defense contractors would buy struggling upstarts for “pennies on the dollar,” but purchasing those startup failures was in the national interest.

He told the audience on a panel that I moderated at the institute: “You’ll have a lot of VCs that lose a lot of money and you get a lot of free [intellectual property] on the back end that keeps us safe.”

What could save some of the investments would be the Department of War changing more quickly how it buys technology. Chip Walter, a general partner at defense investment firm Marlinspike, which has invested in Anduril, told me the Pentagon needs a “two-track approach” to buy traditional military technology both for the “high-end fight” and the low-cost approach many startups are taking.

In the meantime, “we need to stop vilifying the [prime contractors],” said Walter, who was a former director at Northrop Grumman. “We wouldn’t be doing what we’re doing in Iran without the primes.”