The ink is dry on the reauthorization of federal aid for defense startups, and the celebratory backslapping in D.C. is deafening. Every mainstream outlet is running the same tired script: "The government is finally fueling innovation to beat our adversaries."
They are wrong. Dead wrong.
This isn't a fuel injection for the next generation of defense tech. It is a morphine drip for companies that should probably be dead already. By doubling down on the Small Business Innovation Research (SBIR) programs and similar federal lifelines, the government isn't building a "Silicon Valley of Defense." It is building a subsidized playground for "SBIR Mills"—companies that have mastered the art of winning grants without ever shipping a single piece of hardware to a soldier.
I have spent years watching brilliant engineers get trapped in the gravitational pull of government grants. I have seen founders pivot away from solving hard problems because a specific, narrow federal solicitation offered an easier paycheck. This bill doesn't fix the "Valley of Death." It builds a scenic bridge over it that leads nowhere.
The Myth of the Venture Capital Gap
The lazy consensus argues that defense startups need federal aid because venture capital is too scared of the long procurement cycles. The logic goes: if the government doesn't step in, the best tech stays on the shelf.
This ignores how markets actually function. Capital isn't scared of defense; capital is scared of bad customers.
The Pentagon is the world’s most difficult customer. It doesn't matter if you have a $50 million grant if the Program Executive Office (PEO) won't sign a Program of Record. Real venture capital flows toward products that can scale. When the government hands out participation trophies in the form of Phase I and Phase II grants, it artificially inflates the valuations of companies that have no product-market fit.
Imagine a scenario where a software company wins a $1.5 million grant to develop an AI-driven targeting system. They spend two years "innovating" to meet the specific requirements of that grant. By the time they finish, the underlying technology is obsolete, and the actual end-user—the operator in the field—was never even consulted.
This isn't innovation. It's high-stakes bureaucracy.
The Rise of the Professional Grant Hunter
We need to talk about the SBIR Mills. These are firms that exist solely to harvest federal aid. They don't build products. They build proposals.
Data from the Small Business Administration shows a staggering concentration of awards going to a handful of repeat winners. Some firms have won hundreds of grants over decades without ever transitioning a major technology to a commercial or defense program.
By reauthorizing these funds without radical reform to the "transition" requirements, we are effectively taxing the productive economy to fund a permanent class of professional petitioners.
- The Trap: Founders think a grant is a "win."
- The Reality: A grant is a distraction. It forces you to hire compliance officers instead of engineers.
- The Result: You become a consultant for the government rather than a product company.
The Valley of Death is a Feature Not a Bug
The "Valley of Death" refers to the gap between a successful prototype and a production contract. The common wisdom is that we need more "bridge funding" to help startups survive this gap.
I disagree. The Valley of Death is the only thing keeping the defense industrial base from becoming even more bloated than it already is.
In the commercial world, if your product doesn't sell, you go out of business. In the defense world, we’ve decided that every startup with a cool slide deck deserves a "bridge." This prevents the "creative destruction" necessary for progress. We are keeping weak companies alive on life support, which sucks the oxygen (and the talent) away from the few companies that actually have a chance at disrupting the incumbents like Lockheed or Northrop.
If a startup cannot convince a private investor or a specific military branch to put skin in the game after two years of development, the technology is either not ready or not needed. Funding the "gap" just delays the inevitable.
Stop Solving Problems That Don't Exist
The biggest failure of federal aid programs is the "top-down" nature of the requirements. A bureaucrat in a windowless office writes a "Problem Statement." Startups then warp their entire business model to solve that specific, often misunderstood, problem.
This is the opposite of how true disruption happens.
SpaceX didn't wait for a federal grant to tell them that reusable rockets were a good idea. They identified a massive inefficiency and risked everything to solve it. Palantir didn't start by winning a small business set-aside; they built a tool and fought tooth and nail to prove its worth in the field.
When the government provides easy money, it removes the desperation that drives true breakthroughs. It creates a "comfortable" environment for innovation. And as anyone who has ever built anything worth having knows: comfort is the enemy of greatness.
The High Cost of "Free" Money
There is no such thing as "non-dilutive" federal aid. It is the most expensive money a founder will ever take.
While you aren't giving up equity, you are giving up:
- Velocity: The time spent on government reporting and audit compliance is time not spent on product.
- Focus: You start building for the "Contracting Officer" instead of the "End User."
- Optionality: Deeply integrating with government requirements often makes your tech less viable for the commercial market.
I have sat in boardrooms where we had to decide between a $2 million government contract and a $20 million commercial pilot. The "safe" founders choose the government money because it feels stable. That choice is almost always the beginning of the end for their growth. They stop being a tech company and start being a defense contractor.
The Solution Nobody Wants to Hear
If we actually wanted to fix defense innovation, we wouldn't be signing bills to hand out more aid. We would be doing the following:
1. Kill the Grants, Buy the Gear
Stop giving companies money to "try." Start giving them contracts to "do." The government should be an aggressive early adopter, not a venture capitalist. If a startup has a drone that works, the Army should buy 500 of them tomorrow using simplified acquisition authorities. No more three-year "study" phases.
2. Radical Decertification
If a company has won more than five SBIR Phase I awards without a Phase III (commercialization) success, they should be barred from the program for a decade. Clean out the mills and let the hungry founders eat.
3. Open the Testing Ranges
Money isn't the primary bottleneck; access is. Startups can't test electronic warfare gear or long-range munitions in their backyard. Instead of giving a startup $500k, give them free, immediate access to White Sands Missile Range and a team of military technicians.
The Brutal Truth
This reauthorization bill is a win for the lobbyists. It’s a win for the consulting firms that help startups "navigate" the federal landscape. It’s a win for the status quo.
But for the warfighter waiting for a capability that actually works? It’s just more of the same.
We are subsidizing a facade of innovation while our adversaries are focused on scale and speed. We are playing with spreadsheets and "milestones" while the world is moving on.
Stop asking for federal aid. Start building things that are so good the government has no choice but to buy them. If you need a subsidy to survive, you’ve already lost the war.
