Reliance on Dual-Use Technology Is a Trap
As currently executed, the U.S. military’s embrace of dual-use technology will not repel the Russians from Ukraine and it will not deter China from invading Taiwan. Instead, it risks ceding America’s technological advantage to its adversaries.
I have been working at the nexus of national security and technology for many years. Initially, I served as a venture capitalist backing frontier technology companies doing security-relevant work. I then acted as managing director of the Army Venture Capital Corporation, a non-profit organization created by Congress to serve the needs of the Department of Defense. Now I am at a new defense-focused firm. In all these roles, I have repeatedly seen portfolio companies encouraged to build their technologies for the commercial market first and only then slap some green paint on them so that they can begin exploring the U.S. defense market.
This approach is what I call the dual-use hack. It is the idea that the way emerging technology companies should work with the Department of Defense is to build commercial applications first and only then move into defense. Incorporating dual-use technology certainly has its place in constructing the defense ecosystem. Among other things, it helps to mitigate some of the Department of Defense’s first-order challenges in technology adoption, particularly with regard to commercial off-the-shelf technologies. But an overreliance on the commercial-first dual-use hack has generally failed to live up to the hopes laid out almost 30 years ago in the 1995 Dual Use Strategy: A Defense Strategy for Affordable, Leading Edge Technology. This strategy was supposed to ensure that the Defense Department could keep costs low, speed up innovation cycle times, and ensure that America’s industrial base did not atrophy in an environment of falling defense budgets. It has failed at all three of these goals. Costs have skyrocketed, new platform generations take literal generations to release, and America’s defense industrial base has been dangerously eroded in many different sectors.
As if these shortcomings were not enough, an over-dependence on dual-use technology has created several problems on its own. It cedes a growing technical advantage to America’s peer adversaries, discourages startups with cutting-edge intellectual property from building directly for the military’s needs, and provides the Department of Defense with a convenient excuse to avoid or delay real reforms. In short, the dual-use hack has its place, but should no longer be the centerpiece of our national security innovation efforts.
A better solution would enable entrepreneurs to focus on solving defense challenges by making the Department of Defense a better customer. A first step would be to award larger contracts more quickly. The Department of Defense can also communicate better with the entrepreneurial ecosystem regarding current capabilities and problem sets, even if this means sharing more information than they are used to. This communication will serve to lower development risks and make agile, customer-centric development possible. Finally, the Defense Department should provide an incentive system and a clear process that drives successful development projects directly from research and development into acquisitions and sustainment.
What Is Dual Use?
The term dual-use is often used casually to refer to any technology that has applications in both the commercial and defense sectors. However, this definition is too broad to be helpful, as the Department of Defense is one of the world’s largest property managers, residential movers, and apparel buyers — all of which means that just about any product or service is theoretically dual-use. To have a nuanced discussion on the role of dual-use technology in national security, it is helpful to think in terms of four more specific categories.
Commercial/Commercial: This includes technologies which are developed for the commercial market and remain predominately commercial. These are what we think of as “commercial off-the-shelf” technologies. Commercial off-the-shelf technologies are a great place for the Department of Defense to find innovation in software, support functions, and in many of the back-office processes that come with running one of the world’s biggest business enterprises. These solutions are valuable because they are built to exacting industry standards and with substantial private investment, all leading to refined solutions that the Department of Defense can adopt inexpensively and at scale.
Defense/Commercial: These are technologies that originated in defense but have a substantial future commercial market. This category includes many of the dual-use technologies on which Silicon Valley was built but which now make up less and less of the technology ecosystem every year. Technologies being developed right now that arguably fall into this category include hypersonic aviation and advanced methods for positioning, navigation, and timing, such as advanced replacements for the Global Positioning System. These are not the technologies that most venture capitalists are looking to fund, nor are they what most people in defense innovation are talking about under the rubric of dual use.
Defense/Defense: Prime contractors, in turn, traditionally focus on technologies that originate in defense and will largely remain focused on the defense market. This category includes innovative munitions, all the exquisite platforms like fighters, bombers, and ships, as well as the highly specialized applications of certain technologies like AI. This category is only dual use in the most liberal definition of the term and tends to receive the least amount of private venture capital funding.
Commercial/Defense: While all the above categories describe capabilities that are technically dual use, technologies that are commercial first and defense second are usually what the venture capital community and the U.S. defense community are referring to when they use the term. These are technologies that lend themselves to the problematic dual-use hack.
The reason the Department of Defense has increasingly promoted the dual-use approach for Commercial/Defense technologies is because they acknowledge that they themselves are a terrible customer. The logic goes something like this: The Defense Department’s acquisitions process is more likely to kill the early-stage startups they want to work with than they are to drive those companies to scale and success. Given this, startups should develop a commercial product first and successfully scale their business in the commercial markets. Only after they reach sufficient commercial success to survive the byzantine maze of defense acquisitions should they try to adapt their products for the military.
The Problems with Dual Use
The greatest danger of the dual-use hack is that it substantially slows the Department of Defense’s ability to adopt and deploy a new capability. This is because there is an inherent delay built into deploying to the commercial sector first. For example, if you developed AI for swarming drones, the Department of Defense might ultimately be your preferred customer. However, if you need to build and refine your technology for the commercial sector first, you might spend years of your company’s life trying to find the right product-market fit in the entertainment or utility industries. If you find commercial success, you can then hire a government team to begin finding a fit for your swarming AI in the Department of Defense.
In other words, if the country’s best startups are encouraged to treat national security customers as secondary, they will be selling the United States capabilities that are, at best, five to 10 years old rather than cutting edge. Additionally, they will only be able to bring these capabilities to the military if they first successfully find a commercial market. In contrast, China’s system of civil-military fusion does not put a five-to-10-year delay on their adoption of new capabilities — which is why China is five to six times faster than the United States in defense acquisitions.
This dynamic is problematic today and becomes increasingly problematic with each passing year. Technology development cycles are consistently speeding up over time. This means that the annual change in technology is greater each subsequent year. If your AI decision support algorithm is three years behind your adversary’s today it might not matter, but in five years that lag will be magnified. Repeat this across all relevant technology sectors and you get a sense for how much trouble the United States is in. Soon the five-to-10-year capability gap created by America’s reliance on dual-use technology will start to feel like a 20-year technology gap.
And this is not the only problem with a reliance on dual-use technology. A bigger one is that many of today’s national security challenges are simply not amenable to dual-use solutions. This means while all the new startups working in defense are focused on dual-use solutions, a handful of major defense contractors are left as the only ones trying to address more pressing challenges. If policy, preference, and process make it nearly impossible for new entrants to solve purely defense challenges, it institutionalizes a defense industrial base that is dominated by fewer than 10 companies. Startups working in national security are forced to either ignore major national security challenges or develop capabilities specifically for sale to major contractors. In either case, it furthers the much-bemoaned industry concentration Washington suffers from today. By making it difficult for startups to enter and compete in these markets, the Department of Defense reduces competition and creates problems for itself like vendor lock, underinvestment in private research and development, lack of agility, and supply chain vulnerability. If the Department of Defense wants to harness American innovation and regain its technological edge, it should find a way to facilitate defense-first entrepreneurship.
When it was first developed, the dual-use strategy was supposed to facilitate deep integration between commercial enterprise and the Department of Defense — something like Military-Civil Fusion with American Characteristics. But that integration never really happened. Instead, the government supported a proliferation of small-dollar research, development, test, and evaluation appropriations programs to generate commercial dual-use solutions. However, those small-dollar programs were never integrated into major acquisition programs, and there were no processes to coordinate with industry in identifying problems and developing capabilities. Moreover, commercial manufacturing was not integrated into Department of Defense needs. In other words, companies could enter the wide end of the dual-use funnel, but they had nowhere to go at the bottom.
Instead, the dual-use hack has become a coping mechanism. It is an implicit acknowledgement that there are real problems in defense acquisitions, particularly with respect to the point of transition between the research and development side of the Department of Defense and the acquisitions and sustainment side — what is colloquially referred to as the “valley of death.” The focus on dual-use solutions acknowledges those problems and even helps to ameliorate some of them. But precisely because it is a partial solution, it helps avoid solving the real challenges. Rather than trying to become a good customer that can move at the speed of relevance, the Department of Defense encourages the commercial sector to hurry while it moves leisurely along more concerned with process than with the momentum of peer adversaries.
So where does the Department of Defense go from here? The solution starts with addressing the first-order acquisition challenges that make a dual-use crutch enticing. At the same time, the Department should better integrate with industry so that it can accrue the benefits of dual-use solutions without the time-to-adoption penalty and without pinching off the supply of potential defense-first contractors.
As argued by Shands Pickett, Trae Stephens, and Steve Blank, the key to solving first-order acquisition challenges comes down to making the government a better customer. The Department of Defense should move more quickly, picking winners and assigning large procurement contracts to them. It should avoid imposing onerous compliance burdens, focus on software first, and celebrate vendors with high margins who deliver at low cost.
Rebuilding a functional relationship with the entrepreneurial community can begin with two straightforward initiatives. The first is to start a two-way discussion between the Department of Defense and the entrepreneurial community, including venture capital firms and company founders. The second is to connect the current research and development funnel more directly to acquisitions so that startups do not need a commercial-first orientation to survive.
Developing a great product, whether it be for the military or for commercial industry requires mindful collaboration between the developer and the customer. The Department of Defense should be willing to engage with industry in an open and candid manor. China’s system of military-civil fusion allows for seamless coordination between the country’s industry and national security organizations. Beijing has studied the extremely successful relationship the Department of Defense and Silicon Valley shared in the mid-20th century — which developed modern semiconductors and the internet — and copied it. Meanwhile, Washington has built a wall between industry and the Department of Defense. Startups live and die based on customer feedback, but Washington asks that companies building for the national security community do so without a clear understanding of its current problem sets and capabilities. U.S. startups are forced to find product-market fit in the dark, which drives up costs and failure rates. Trusting a subset of the U.S. entrepreneurial ecosystem to develop next-generation defense capabilities requires trusting them to help shape requirements, develop concepts of operations, and refine capabilities directly to address real world problems. If that kind of cooperation requires security clearance, the government should find a quick and efficient means for making that happen at scale.
The Small Business Innovation Research grant program is one of the most common ways startups receive funding from the Department of Defense. These grants are designed to help early-stage technologies proceed from initial development up to the prototyping stage. In my experience, the founders of companies receiving these grants — particularly those receiving larger grants under the program — are almost universally convinced this represents their entry point to the defense market. They believe they are only a couple of years away from selling to the Department of Defense at scale. 99 percent of them are wrong. When companies in this program fall into the “valley of death” and fail to break into the defense market, they often permanently pivot away from serving a national security need or simply go out of business. Anecdotally, the most successful founders I have met who are building for national security eventually quit working with research and engineering organizations and go find an acquisitions program manager on their own. This initiative, while laudable, should not be required. Successful research and engineering programs should directly feed into acquisitions. A system that seamlessly moved from capability development to capability adoption would enable founders to build for defense customers first without fear of being swallowed up by the “valley of death.”
I have dedicated my career to supporting the defense innovation base, and I firmly believe that dual-use technology has a role in meeting the military’s technology needs. Some companies should build for the commercial market first, but the Department of Defense cannot always afford to wait for the dual-use track to play out. Companies should be able to build directly for the Department of Defense rather than treating national security as a side hustle. There is no lack of patriotic U.S. entrepreneurs eager to contribute to the country’s defense. Showing industry that developing technology for the military can be successful will cause both financial and human capital to flow into the national security space.
Jake Chapman has been in and around the venture industry for almost 20 years as a lawyer, three-time founder of successful companies, and venture capitalist, including as managing director of the Army Venture Capital Corporation. Jake is the managing director of Marque Ventures, a venture capital firm focused on national security technology. Jake writes infrequently at The Labyrinth and all too frequently on Twitter as @vc.
Image by NATO