It’s a strange time for the defense technology sector. Despite the headlines of record capital, valuations, and defense budgets, there is a growing unease among investors and leadership teams. Production contracts remain as elusive as ever, and a generation of startups is struggling to extend their runways across the valley of death, risking not just those businesses but capital flight from the broader sector. There is little point in trying to decide which one of these diverging narratives is real, or if a defense technology market collapse is imminent. To understand what’s really happening, we should recognize that things are getting Dickensian.

In A Tale of Two Cities, Dickens described the winners and losers in the years before the French Revolution. In London, life was outwardly stable, but risks were ever-present. In Paris, chaos abounded, but so too did opportunity.

Defense technology has evolved into two distinct markets. London is the land of production contracts and annually recurring revenue. Paris is the ever-churning metropolis of prototypes, pilots, and discretionary funding. Anyone assessing the health of the sector should view each market separately. Early-stage, prototype-heavy startups still present opportunity, but are under pressure. Significant new capital is concentrating in production-heavy startups, but risks abound. Production contracts are key to escaping Paris and thriving in London. In their absence, startups will struggle, capital will get exhausted without replenishment, and everyone will need to prepare for mergers and acquisitions.

Sign Up for Our Newsletter

It is the Best of Times, it is the Worst of Times

The best of times are easy to see. The United States creeps closer to a trillion-dollar defense budget. Around $49 billion in venture investment was allocated to defense technology in 2025. The Silicon Valley Defense Group’s top 100 startups have raised over $118 billion. Valuations continue to break records. Palantir is now valued at over $400 billion, greater than Lockheed Martin, Northrop Grumman, and General Dynamics combined.

That success is important, hard-won, but obscures the scarcity and delays plaguing most startups.

The Defense Department obligates over $450 billion annually in contracts, and in 2024, the top five primes received 30 percent of those obligations. Silicon Valley Defense Group’s analysis shows that their top one hundred startups combined secured $4.3 billion in awards in 2025. Government auditors reviewed 450 Defense Innovation Unit prototype other transaction authority awards from 2016 to 2023 and found that only 14 percent converted from prototype to production (although the number rises to 51 percent among completed prototypes). The Pentagon reported only $2 billion of production other transactions across the department in 2024 and does not track the conversion of prototype other transactions into standard contracts. That asymmetry is longstanding, but the volume of startups and investors betting it will change has never been higher.

These two realities will continue to coexist. Fears about the health of defense technology investing have been raised for a number of years, but have yet to materialize and won’t do so in a highly visible manner: private capital allocators don’t advertise failures, investing primarily private capital means there are limited taxpayer dollars at risk to trigger government oversight, and the Defense Department will keep advertising its “innovation” irrespective of reality. Instead, those who care about these markets should assess the health of businesses and capital in each market, respectively. The key question for investors and startups alike is which reality applies to them.

Life in Production

Scaling startups to programs of record, funding locked into the Future Years Defense Program, with clear renewals planned on predictable sustainment timelines has always been the goal of founders. But life in production still contains risks. Renewals are not guaranteed, funding is under continual pressure, traditional competitors can influence Congress, and the political environment beyond the midterms and into the next administration is both unclear and consequential.

At present, only a small number of new entrants are enjoying the apparent tranquility of production at venture scale. Palantir’s Maven and Army enterprise agreements, Anduril’s Army contract and Collaborative Combat Aircraft, Saronic’s Corsair, Hawkeye 360’s internationally diversified contracts show what’s possible.

And in this new defense tech environment, there is also a price attached, Anduril is valued at roughly 30 times revenue, Palantir has traded even higher. The expectations of growth pricing can easily run afoul of Defense Department program delays, with major consequences for all.

Life in Prototypes

Living in prototypes is chaotic and uncertain. What is presented to boards as annual recurring revenue is often cobbled together research, development, test and evaluation dollars, pilot budgets, and innovation funding secured opportunistically by passionate advocates inside the Defense Department and the services. These are the types of discretionary funds constantly at risk of sweeps and reallocation to current operations, and from continuing resolutions and the rules that prohibit new starts.

Prototype-heavy startups are also where the venture investor base is under most pressure. Long-standing defense specialist investors are nearing the end of fund lifecycles and their own patience, while generalist venture capital firms that entered defense as part of a trend are now discovering that defense procurement does not scale like consumer software and does not offer clear exit paths.

But early-stage startups also represent opportunity. The same operational pressures that squeeze discretionary budgets create demand in new areas. Interceptors and long-range munition stockpiles are dwindling, traditional approaches will take too long to replenish, and as a result startups like Castelion and Covenant are emerging with initial contracts and real capital presenting huge opportunities to operators and investors. The question is whether those contracts convert into programmed dollars at scale.

Watch Investors More than Startups

There will not be a single event that will signal a turn into chaos or revolution. All that is required is for current conditions to continue, and math will do the rest over time.

The dynamic between aggressively optimistic startups and a sluggish Pentagon has remained constant for years. The key difference today is the capital, meaning the behavior of investors is a better predictor for increased growth, churn, or a marked downturn. In that context, we should be more concerned about earlier-stage startups trying to cross the valley of death. Numbers don’t get advertised publicly, but the general lifecycle of venture funds tells us that the investors that fueled raises in the early 2020s should now have fully deployed their capital and are nearing the age where limited partners expect returns, in the form of cash from exits. They will not have the patience or capital to keep funding portfolio companies short on liquidity and, without production contracts, growth investors will not provide fresh runway.

Exposure runs through both markets but in markedly different ways. For the prototype heavy market, it concentrates in companies whose path to production and real revenue depends on a single champion, in oversupplied categories — think attritable drones, counter-drone or generic AI — in companies backed primarily by generalist funds that are investing on a trend rather than deep experience, and in pure play software providers that don’t yet own workflows. The early-stage market is also at greater risk of capital flight, as evidenced by slowing seed and Series A deals.

The production market exposure shows a different profile. Companies are more concentrated and similar in overall profile with production contracts, high capital intensity, hardware and software enabled by AI, so differences are in business performance and the associated implications. Lumpy revenue misses from program delays can lead to liquidity issues and painful capital solutions, technology failures or operational issues create category concerns that don’t apply to traditional primes, and multiple compressions could mean challenging raises even if bookings are growing. And all the while traditional primes capitalize on new innovations with updates to traditional solutions.

Much of the recent spike in defense technology investment has accrued to this smaller group of companies, such as Hadrian, Apex, and Neros, many of whom are experiencing significant market momentum even if true financial success is difficult to determine. Fresh capital means more time for investors and, along with real production contracts, provides significant financial resources to cushion department delays. But the capital intensity of these businesses is also extreme, and growth investors are far less forgiving than early-stage investors if things go off track. The biggest question for startups in production is how to generate meaningful returns for investors. Current valuations make acquisitions by the traditional defense industry extremely challenging and the initial public offering market for defense specialist businesses is decidedly unclear due to a lack of visibility on future revenue and earnings. Hawkeye 360’s recent listing was oversubscribed and successful at opening, but the stock has declined since despite strong business execution.

Both markets will feel further pressure in the coming years that could worsen conditions or precipitate capital flight: another impending continuing resolution, potential Congressional changes from midterm elections, operational tempo draining discretionary dollars, or a change of administration. These government events will add to the strain on investors, particularly early-stage, as funds from the early 2020s have fully deployed capital, try to raise new funds, and startups seek new capital.

Winners and Losers

The market-level failure modes are different for prototypes and production. Prototype-dependent companies will quietly run out of money, and no single failure will count as news. Market-level issues will only be discernible when new capital stops arriving to bet on the next startup or round. In this way, the fall of the prototype market can only be truly recognized after it has occurred. Everything before that point is guessing about the conviction level of investors.

For production startups, individual failures will be more noticeable. A small number of high-profile collapses could more easily dissuade investors from continuing to allocate capital. Again, the key determinant will be investor conviction, and that will only become obvious after the fact. With the recent influx of production concentrated capital, it will be years before that assessment can be made.

The central challenge in assessing the health of the defense technology sector is that it requires watching what doesn’t happen. Funds that were expected to close but didn’t, financings that don’t happen or occur with debt and restrictive terms, companies that go silent, and contracts that aren’t awarded. Tracking aggregate awards and major wins, especially for production contracts, is the best leading indicator of the health of the sector. The least helpful thing to watch is policy and budget announcements from either the executive or legislative branches.

Four rough categories are emerging for the winners and losers across both markets.

The Deserved Few

The leaders of this group are already clear, even if they too will experience duress in the future. Corrections and chaos create opportunity, effective operators can still find paths to production at scale, and these leaders will still have access to necessary capital.

The Deserved Losers

Most regular startups don’t survive — and not every defense tech startup deserves to either. The recent flood of capital has enabled startups that were never going to scale, even if the Defense Department were a more reliable customer. Not all of the current uncrewed systems, counter-drone, hypersonic, cyber, or AI startups are required, especially if the technology is generic, the leadership inexperienced, or the pitch is only PowerPoint deep. The crowd will be thinned in multiple ways: down selects, capital exhaustion, and acquisitions.

The Undeserved Losers

The recent enthusiasm for defense technology has created real technology and great companies that solve vital operational problems. Right now, they are racing the clock through meaningful pilots, requirements processes, budget justifications, and authority to operate issues. Even if they do everything right, they may find themselves out of cash, before necessary contracts arrive, in a capital environment with insufficient conviction to fund another six to twelve months of runway. Others may collapse under prior valuations and liquidity preferences in a capital environment that has turned skeptical.

Opportunistic Acquirers

This category receives the least attention and may be the most interesting. Traditional primes dominate production but historically execute acquisitions based on clinical assessments of the earnings before interest, taxes, depreciation, and amortization value of production contracts. If primes and private equity platforms can broaden their thinking, distressed startups with real contracts, differentiated technology, but limited runway could present a new vector for acquisition as valuations tumble. Venture-funded technology, talent, and sub-scale contracts could join platforms that provide production scale, bigger technology stacks, contract vehicles, and the stability of traditional sales and government affairs. If executed well, the Defense Department could get venture-funded technology into production through traditional business models. While beneficial for the Defense Department and primes, venture investors would be the billpayer further undermining the argument to allocate future capital.

What to Do

Despite all this uncertainty and complexity, the necessary action for everyone is clear. Focus maniacally on clear assessments of the viability of production contracts. That is, contracts have a buyer with procurement dollars rather than a champion with an innovation budget, a procurement line item in the appropriated budget, and a contract vehicle that allows awards without further competition. Exercises, competitions, pilots, and prototypes are great, but don’t guarantee production at scale.

This is also the moment for everyone to start thinking seriously about mergers and acquisitions, though the vision startups sell doesn’t align with the contracts primes buy. Founders are eternal optimists, and one contract can change everything. But, with funding less certain, unprofitable startups should be looking at exits alongside financing with, at minimum, six months of runway. Defense industry acquisitions take time, and a discounted landing beats insolvency that strands the team and technology. Meanwhile, primes and private equity platforms should be developing disciplined roll-up strategies to play the clock and acquire distressed assets at the right price.

The tension between these incentives — sell early, buy late — shows the paradox the defense technology sector is experiencing. The operational needs, new technology, and new capital driving the current boom are real. But so too is the dearth of production contracts required to sustain that boom.

If the Defense Department and the services are truly concerned about the health of the defense technology sector and associated capital markets, all they need to do is award production contracts at scale. No additional policy reforms, initiatives, or organizational changes are required. In the absence of material changes to contract volume and value, the fortunes of both startup markets will continue on their current trajectories.

Market turbulence is no reason to abandon the defense technology sector. The United States will continue to need venture-backed startups to meet the threats of today and tomorrow. Smart investors and operators who find pathways to production, and carefully manage their capital, will still be able to deliver capability for the nation and real returns to their limited partners. Opportunity will abound even if it’s with less fanfare.

Over the coming years, defense technology will continue to have two startup markets. Which companies end up succeeding is not yet decided, and we will remain in the best and worst of times.

Write for Cogs of War

Ben FitzGerald is the executive chairman of RAIC Labs and serves on the boards of Rebellion Defense, Morning Consult, percipient.ai, and UBIQ Aerospace. He was previously the CEO of Rebellion Defense, Partner for dual-use investing at Lupa Systems, executive director for strategy, data, and design in the Office of the Under Secretary of Defense for Acquisition and Sustainment, a professional staff member on the Senate Armed Services Committee, and director of the Technology and National Security Program at the Center for a New American Security. The views expressed are his own.

**Please note, as a matter of house style, War on the Rocks will not use a different name for the U.S. Department of Defense until and unless the name is changed by statute by the U.S. Congress.

Image: Midjourney