Katherine Boyle on the Army's transformation initiative and what's next for American Dynamism investing

Summary

Katherine Boyle, general partner at Andreessen Horowitz, argues that American Dynamism is roughly three to four years into what she sees as a 30-year project — and that the cultural win of making defense tech mainstream in venture is not the same as the mission being accomplished.

Army Transformation Initiative

The week's biggest under-reported story, in Boyle's view, is the Army's Transformation Initiative announced by Secretary Driscoll and General George. The Army went public — including an appearance on Fox & Friends — to declare it would divest obsolete platforms, cut civilian roles it no longer needs, and compete programs on shorter cycles rather than locking into decade-long contracts. Boyle's read is that DOGE gave political air cover to reformers inside the DoD who had been pushing for acquisition changes for years without traction. The Humvee is her sharpest illustration: developed in 1980, declared inadequate for IED warfare in 2004, and still in production in 2025.

The procurement model she describes is the structural problem. The DoD has historically required a decision on a platform that would receive no updates and could not be cancelled — a constraint no private-sector CEO would accept. The Transformation Initiative signals the Army wants to recompete programs every year or two as technology changes, which structurally opens the door for startups that couldn't previously compete against entrenched multi-decade contract holders.

What the market map looks like

Boyle pushes back against the instinct to predict TAM in this category. Early skeptics dismissed Andreessen Horowitz's defense portfolio on the grounds that selling to DHS was a small market — a view she describes as comical in retrospect. Her point is that the markets themselves grow as the national priority shifts.

For founders trying to find uncrowded ground, she points to two directions. First, genuinely unglamorous corners of defense — logistics, supply chain, component manufacturing — that no existing prime or startup is prioritizing. Second, the tier-one supplier model: companies that position themselves as partners to Anduril or SpaceX rather than competitors. She notes that U.S. drone motor manufacturing is almost entirely absent domestically, with production concentrated in China, as one example of a gap that is real but may not fit the venture power-law model cleanly.

On the SpaceX alumni network specifically, Boyle says an a16z analysis found hundreds of companies formed by ex-SpaceX founders across wildly different sectors over the past decade — including Radiant Nuclear and Castelion, which is building hypersonic weapons and is in the a16z portfolio.

M&A and the prime consolidation thesis

Boyle's more contrarian prediction is that the legacy primes — Lockheed, Boeing, Raytheon, and the rest of the big five — will be forced into acquisitions they have historically avoided. The primes have almost exclusively acquired companies that never took venture dollars, steering clear of the startups seen as bleeding-edge competitors. She thinks that changes within five to ten years as DoD competition intensifies and the primes can no longer recruit the engineers or build the R&D in-house.

She raises the possibility of another "Last Supper" moment — a reference to the 1990s episode when the government effectively pressured primes into forced consolidation as defense budgets contracted. Her version runs in the opposite direction: a wave of acquisitions driven by the need to absorb startup capabilities rather than shrink the supplier base.

For earlier-stage companies, she sees M&A already being used tactically — acquiring a specific capability to qualify for a major prime contract — in a way that wasn't common a few years ago.

Investing posture on geopolitics and timelines

Boyle says her bias as an early-stage investor is to ignore near-term headlines and hold to the long cycle. The a16z investment in Hadrian, the precision manufacturing company, was made when re-industrialization was an early signal rather than a policy consensus. Her frame is that globalization took decades to peak and the pendulum swinging back will also take decades — meaning the generational companies in this wave are still being founded now.

The platform build-out

A16z has built a Washington office staffed across both parties, with specialists covering different DoD branches. Boyle's partner Leila runs go-to-market in DC and has published a public glossary of DoD contracting terminology and acronyms. The thesis is that defense and public-safety companies need a fundamentally different kind of support than enterprise SaaS — including understanding the right contracting vehicles, the right people on the appropriations committee, and how to navigate BD with government buyers — support that a16z is trying to provide at the platform level rather than leaving each portfolio company to figure it out alone. Flock Safety's "second city" strategy of surrounding major metro areas by winning smaller surrounding municipalities is one playbook Boyle cites as a model others are now replicating.