Visit tech startup Skydio’s headquarters on the San Francisco peninsula in California and you’re likely to find flying robots buzzing on the roof overhead. Docking stations with motorised covers open to allow small drones that resemble the TIE fighters from Star Wars films to take off; when each drone lands back again, they close. The drones can fly completely autonomously and without GPS, taking in data from onboard cameras and using AI to execute programmed missions and avoid obstacles.
Skydio, with more than $740m in venture capital funding and a valuation of about $2.5bn, makes drones for the military along with civilian organisations such as police forces and utility companies. The company moved away from the consumer market in 2020 and is now the largest US drone maker. Military uses touted on its website include gaining situational awareness on the battlefield and autonomously patrolling bases.
Just across the water from its headquarters is Skydio’s manufacturing facility, where about 200 workers assemble hundreds of drones a month, including the defence model. The company has agreements to supply its drones to the US defence department, as well as 24 US allies, and they have been used by the Ukrainian military fighting Russia. “It is an absolute certainty that small, inexpensive, software-defined systems with rapid iteration are the future of defence,” says Adam Bry, Skydio’s co-founder and chief executive.
Skydio is one of a number of new military technology unicorns – venture capital-backed startups valued at more than $1bn – many led by young men aiming to transform the US and its allies’ military capabilities with advanced technology, be it straight-up software or software-imbued hardware. The rise of startups doing defence tech is a “big trend”, says Cynthia Cook, a defence expert at the Center for Strategic and International Studies, a Washington-based-thinktank. She likens it to a contagion – and the bug is going around.
According to financial data company PitchBook, investors funnelled nearly $155bn globally into defence tech startups between 2021 and 2024, up from $58bn over the previous four years. The US has more than 1,000 venture capital-backed companies working on “smarter, faster and cheaper” defence, says Dale Swartz from consultancy McKinsey, adding that Europe has seen an uptick in defence tech startups too. While most of the funding has gone to US-based companies, some, such as German startup Helsing, have seen significant amounts. Yet a sector set on reinventing defence with Silicon Valley values also raises concerns, including whether it could bring us closer to war – and Donald Trump looms large.
As the upstart defence industry sees it, the current system is not set up to meet the needs of the modern war fighter. We are entering a new era where machines go to war, albeit working with humans, and there is a huge need for autonomy and AI that the “defence primes” – the massive companies the defence department has traditionally partnered with to build ships, planes, tanks and strategic deterrence weaponry such as Lockheed Martin, RTX and Boeing – do not have the right muscles to deliver. As a result, the US risks losing its edge in its ability to respond, which is something the startups say they can help fix.
And the potential rewards are enormous. The US spends about $850bn annually on its military, approximately half of which goes on procuring new items or maintaining old equipment, while the total military spending globally is more than $2.4tn – an amount set to rise significantly as Europe assumes an increased burden for its own security.
Not, says the defence startup sector, that it is just about money. Imbuing it is a zeal to help the US and its allies retain a military advantage over their adversaries in an increasingly dangerous world. “The engineering elite of Silicon Valley has an affirmative obligation to participate in the defence of the nation,” states the preface of The Technological Republic, a new book by Alex Karp, which can be viewed as a manifesto for the fast-rising industry. Karp is the chief executive and co-founder, along with billionaire Peter Thiel, of AI-driven software company Palantir Technologies, which, with Elon Musk’s SpaceX, is seen as a trailblazer for the upstart industry.
And the sector is bullish about its prospects under the Trump administration, which has signalled it wants to revamp and modernise procurement. Their approach, say the startups, can deliver more for less money.
The types of technologies the defence upstarts are working on are many and varied, though autonomy and AI feature heavily. To give a flavour, in addition to autonomous aerial drones such as Skydio’s, there are those that travel on the surface of the sea and underwater, as well as generative AI to enhance military planning and decision making, AI-powered counter-drone technology, autonomous strike weapons and even AI pilots for fighter jets, negating the need for human ones.
One startup known for its aggressive and rapid expansion is southern California-based Anduril Industries. The company, which has received $3.7bn in venture capital funding and is valued at $14bn, was co-founded in 2017 by the inventor of the Oculus Rift virtual reality headset, Palmer Luckey. The business, which is focused on autonomous systems and weapons – including the Thunderbird-like Fury war fighter – is building a manufacturing facility in Ohio and is also reportedly planning a drone factory in the UK to serve as a European base.
It is only in the past few years that defence has come to be seen as a viable market for startups. The prevailing belief before that was the US defence department was limited to working with the primes.
Yet the lion’s share of the funding is still going to the primes, the startups complain. The industry is “emerging fast in the outside world, but not so fast in terms of budgetary reassignment”, notes Andy Lowery, chief executive of Epirus, another high-valuation startup that is focused on disabling swarms of many thousands of attack drones using high-energy microwave forcefields, and is working with the British military’s Army Futures directorate on how the technology may be useful to the UK.
The problem – and what needs to be disrupted, many in the emerging sector argue – is the US defence department’s antiquated system of military acquisition and procurement, which has long budget planning cycles and is oriented towards buying large, expensive, static hardware systems that can take many years to come to fruition and leave little space for innovation because the contracts are so overly prescriptive.
The defence department is “very stuck in old cold war ways of doing things”, says Michael Brown, a partner at venture company Shield Capital and the former director of the US military’s defence innovation unit (DIU). And that strategy does not work well when threats are changing rapidly and new technology needs to be leveraged.
Ryan Tseng, co-founder and chief executive of Shield AI, a defence tech unicorn focusing on AI pilots, agrees that military acquisition is overdue a shake-up. Tseng would like to see “millions of AI pilots”, so the US and its allies are ready for the future of war. “[But] at the moment it is basically at zero adoption,” he says.
Another defence tech startup Castelion, also based in southern California, was founded in 2022 by former SpaceX employees and is pursuing autonomous strike weapons: hypersonic long-range missiles with AI capabilities. Hypersonic weapons – of which China has the world’s leading arsenal – can travel at above five times the speed of sound and though the US doesn’t have any yet, it is actively developing them with some of the primes. Castelion has won military contracts to build and test its prototypes, and in January announced its first big funding round of $100m to accelerate test cycles and build production facilities.
Castelion’s approach is different from what the US is known for, says Bryon Hargis, its co-founder and chief executive. Rather than build limited numbers of costly and long-lasting systems with exquisite, high-end capabilities, Castelion wants to manufacture fear-inspiring quantities of low-cost “sufficiently capable weapons” – an approach, he says, that will “actually achieve a deterrent effect”. It is hoping to deliver its first weapons at scale in 2027, which would be exceedingly fast.
Yet the push to remake defence has left some worried. Elke Schwarz, a professor of political theory at Queen Mary University of London, has analysed the effect of venture capital dynamics on military norms. It is not, she argues, that the defence sector may not benefit from an overhaul – a shift towards new technologies and greater agility. But the high risk/reward venture capital funding model comes with such huge expectations of rapid growth – the company must scale up fast to try to reach a high valuation – it can lead to products being oversold.
“The hyperbolic AI discourse that we know from the domestic sector also happens in the military environment,” she says. The risk is that the products fielded, which often rely on hastily produced prototypes being quickly tested and improved, do not work as advertised and are flawed and unreliable in their capabilities. (Though, to be sure, such potential problems are not just limited to startups, as demonstrated by the issues that have dogged the Lockheed-manufactured F-35 fighter jet). Schwarz is particularly concerned that the acquisition of all these relatively low-cost tech products pushes the US and its allies closer to wanting to use them in war.
The startups counter that an agile, iterative approach does not mean they are delivering inferior products, and checks and balances at the defence department ensure the technologies that are adopted are up to snuff. Broader acquisitions are not made until the technology has been put into the hands of operators to test, said a DIU spokesperson: “They provide the unvarnished opinion on whether [it] works as promised, addresses defence problem sets and would be useful in key scenarios.”
On pushing us towards war, the companies note that they do not decide how what they sell will be employed. What, if any, payloads the military attaches to Skydio’s drones is up to the military. “These things are just tools,” says Castelion boss Hargis. “Our intention is just to give the US and its allies the best possible capabilities… [and] without military strength, I truly believe we’re all in for a worse future.”
Moving forward, doubtless the primes will do some adapting – indeed some have started up their own venture capital funds. None of the primes approached or their US trade associations agreed to be part of this feature, though Kevin Craven, the chief executive of the ADS Group, the UK trade association for the aerospace, defence, security and space industries, highlights the longevity and experience the primes bring. “[They] are innovators in their own right,” he says.
It is also anticipated that, as the new military tech industry matures and takes shape, there will be some consolidation – even among the larger players. Lowery points to the way the primes rose, which was through mergers, particularly after the cold war. “History repeats itself,” he says.
Roberto González, a professor of cultural anthropology at San José State University, California, who has also dissected how the tech sector is transforming the US military-industrial complex, adds that it is worth thinking too about what may not get funded when venture capital money is ploughed into defence; for example, some climate startups are now pivoting their pitches to defence. “There are opportunities lost through military spending,” he says. “Opportunities to invest in biomedicine, clean energy, smart farming technologies and much more.”
Meanwhile, there is “a little pent-up demand” to see if the startups that have been funded are successful, says Shield Capital’s Brown. If companies “go under” and returns are not realised, it could mean an “investment winter” for the technology, which “wouldn’t be good” for competition. For Brown, it all hinges on reforming the US defence department’s acquisition and budgeting system.
And whether that will be as bold as the interlopers are hoping under Trump remains to be seen.
