Frank Kendall argues that criticism of defense primes is misplaced. The defense industrial base builds what its customer, the Department of Defense, asks for. To get cheaper, simpler, and more innovative products, the services must change their requirements and demand them. The problem lies with the customer, not the supplier.

In a major strategic shift, the Pentagon is asking prime defense contractors to invest their own capital—billions of dollars—to expand munition production "on spec." This pushes immense financial risk onto publicly traded companies, a difficult ask given the government's historically cyclical and unreliable purchasing patterns.

The government's procurement process often defaults to bidding out projects to established players like Lockheed Martin, even if a startup presents a breakthrough. Success requires navigating this bureaucratic reality, not just superior engineering.

Unlike traditional contractors paid for hours, Anduril invests its own capital to build products it believes the government needs. This model incentivizes speed and effectiveness, as profit is tied to successful products, not billable hours. This shifts the financial risk from the taxpayer to the company.

Unlike traditional contractors paid for time and materials, Anduril invests its own capital to develop products first. This 'defense product company' model aligns incentives with the government's need for speed and effectiveness, as profits are tied to rapid, successful delivery, not prolonged development cycles.

To minimize risk, government contracts often require bidders to have prior experience building the exact same system. This seemingly prudent rule creates a catch-22, barring new entrants and locking in a small number of incumbents who can then dominate the market and inflate prices.

Today's cumbersome defense acquisition system was designed to solve the 1980s problem of fraud, like overpriced screwdrivers. While successful at preventing that "Type 1 error," it created a massive "Type 2 error": an inability to procure technology at a relevant speed.

The defense procurement system was built when technology platforms lasted for decades, prioritizing getting it perfect over getting it fast. This risk-averse model is now a liability in an era of rapid innovation, as it stifles the experimentation and failure necessary for speed.

Traditional defense primes are coupled to customer requirements and won't self-fund speculative projects. "Neo primes" like Epirus operate like product companies, investing their own capital to address military capability gaps, proving out new technologies, and then selling the finished solution.