Why You Can’t Call in an Air Strike with an iPhone
Christian Brose, The Kill Chain: Defending America in the Future of High-Tech Warfare (Hachette Books, 2020)
Between 1996 and 2011, the U.S. military spent $6 billion to develop and field a new tactical radio. Even in the context of U.S. military spending, $6 billion is a big chunk of change. By comparison, the Air Force spent approximately $3 billion to develop and procure the pathbreaking MQ-1 Predator. The Predator ushered in a new era of drone warfare, whereas the tactical radio project was cancelled before it could produce a single radio.
Harris Communications was one of the companies that hoped to win the contract for this program. The leaders at Harris foresaw the enormous technological and program management challenges that awaited the winner. Harris didn’t win the contract, but that didn’t stop the company from taking advantage of the opportunity. It invested $200 million of its own research and development dollars to develop a radio system with less ambitious performance goals and that would be unencumbered by unwieldy Pentagon acquisition regulations. Harris succeeded in 2008 with the PRC-117G radio, which could support a modest tactical voice and data network that has since become the workhorse standard for the Army and Marine Corps.
Harris’ radio has many of the hallmarks in which proponents of greater commercial technology in the U.S. military believe. Namely, nimble commercial firms can be more effective if they are less constrained by the Pentagon’s cumbersome acquisition bureaucracy. Although Harris is not a Silicon Valley start-up, the success of its radio when compared to the more ambitious radio-that-never-was illustrates the problems of the acquisition system and highlights the attractiveness of letting technologists have more freedom to work.
The discourse about emerging technology in the Department of Defense today is centered around the military potential of commercially developed information technology. If anything, the role of commercial technology is clearer today than it was in in 1996. Pivotal, a software company, worked with the Air Force’s Kessel Run program to transform tanker refueling schedules with easy-to-use software. Artificial intelligence-driven drones have the potential to overwhelm defenses while quantum sensors can detect even stealthy submarines beneath the waves by their minute gravitational signatures on the wavetops. The military potential for these and other commercially developed technologies is substantial.
For many technologists eager to help the U.S. military though, the conversation is often tinged with a sense of frustration that the military does not adopt commercial technology more readily than they believe it should. This frustration often focuses on the role of the Department of Defense’s acquisition bureaucracy. Congress, for instance, has asked pointed questions to the Army about why it was reluctant to adopt the commercially developed Palantir intelligence analysis software system. And, former Google CEO Eric Schmidt once proclaimed to the head of U.S. Special Operations Command that “If I got under your tent for a day, I could solve most of your problems.” Tesla and SpaceX CEO Elon Musk confidently stated earlier this year that a manned F-35 would be no match against a semi-autonomous drone in air-to-air combat. There is a strong sense among interested technologists that breakthroughs in the commercial sector will be critical to warfare in the future and that the overly restrictive Pentagon processes and stodgy culture are impediments to that future.
The frustrations of commercial technologists should concern the Department of Defense. The under secretary of defense for research and engineering’s modernization priorities include artificial intelligence, biotechnology, and other technologies where the commercial sector is leading development efforts. It is clear that commercial technology companies will be an important part of an expanded defense industrial base, giving weight to technologists’ concerns. Some firms may find the defense sector to be an economically challenging market; a reputation for frustrating red tape may make it even less attractive. Most concerning, though, is that simple frustrations about Pentagon bureaucracy are an easy conclusion to draw that offers little hope about whether the situation will improve. Such a conclusion obscures deeper exploration into the reasons why commercial technology is not more readily adopted by the military.
Is “The System” the Only Obstacle?
There is no shortage of criticism of the defense acquisition bureaucracy, but is that the only reason why troops aren’t calling in air strikes from iPhones and using artificial intelligence to control drone swarms? Two other reasons might also be considered: first, adapting commercial technology for military purposes is harder than it seems; and second, the military might not be fully convinced that available commercial technologies are what it wants.
Christian Brose’s new book, The Kill Chain: Defending America in the Future of High-Tech Warfare, is an insightful analysis of the bureaucratic obstacles to the adoption of commercial technologies by the U.S. military. His years of experience on Capitol Hill imbue his book with a sense of context that moves the conversation past the frustrations that technologists have expressed. He deftly describes the bureaucratic and political power structures and incentives that keep the U.S. military from more readily integrating commercial technology. It is a powerful contribution to the conversation about technology and defense.
Brose’s critiques are more nuanced than those of many frustrated technologists. However, he still confines his arguments to issues about the political incentive structure and acquisition bureaucracy. To keep advancing the conversation, we should consider these two possible obstacles along with Brose’s critique of the bureaucracy.
The Bureaucracy Is Imposing Obstacles
In his book, Brose argues that the Pentagon’s organization, process, and incentives are preventing commercial technology from taking root in the military. He argues that commercial information technologies such as artificial intelligence will define the future of conflict and that the United States is underinvesting — both financially and organizationally — in those technologies. Meanwhile, Brose argues, America’s adversaries have watched, learned, and stolen a march on new technologies, including artificial intelligence, quantum computing, biotechnology, and space systems.
Brose offers a well-thought-out diagnosis of why this underinvestment exists, even though the United States correctly envisioned the role that commercial information technologies would play as far back as the early 1990s. America’s hubris about its supremacy made it slow to act, he argues, as did a two-decade counter-insurgency and counter-terrorism odyssey that distracted the United States from making progress. Brose further argues that the Pentagon is incentivized to value stakeholder consensus over decisiveness, with a budgeting process that favors incumbent programs over new ones and an acquisition system that favors process compliance over effective outcomes. The result, he believes, is a defense establishment that is unable to change course until it is too late.
Brose’s observations and arguments about the organizational hurdles to greater commercial technology adoption by the U.S. military force us to reflect on the values for which the acquisition bureaucracy strives. For instance, his analysis of the acquisition system’s prodigious regulatory burdens, which exist to ensure fair competition and save money, forces readers to question the purpose of all the red tape: Is saving pennies worth the trouble when the future of U.S. national security is at stake? Brose believes that, when it comes to confronting “emerging great powers with chips on their shoulders and serious military technology ambitions,” the United States has done what it did during the Cold War when it “pick[ed] winners … the people who could succeed where others could not, and the industrialists who could quickly build amazing technology that worked. Other concerns, such as fairness and efficiency, were of secondary importance.”
However, there is something to be said for fairness and efficiency. Done right, fair competition yields a diversity of approaches that is more likely to prepare United States to endure the shocks and surprises of clever and adaptive adversaries. Even ballistic missile pioneer Bernard Schriever — one of Brose’s picked “winners” — hedged his bets by pursuing multiple approaches that yielded the Atlas and Titan missiles. Cost-effectiveness is also underrated. America’s national resources are finite. And, the ongoing COVID-19 pandemic is only one example that should encourage reflection on budget priorities. Brose correctly diagnoses the ills of the defense acquisition bureaucracy, but its goals are still worthwhile. Brose is right that mindless adherence to acquisition rules without considering the wider context wastes time and effort. One might be better served by continuing the hard work of reforming the bureaucracy — not sidestepping it.
Defense Technology Is Harder Than It Looks
Another reason why emerging commercial technologies may not be more readily adopted by the Pentagon is that adapting such technology for military use may be harder than it seems. This chance is a distinct possibility. Maaike Verbruggen argues that military expectations for artificial intelligence should be tempered. Artificial intelligence is not yet capable of performing subjective tasks where judgment is required; for instance, it still struggles to accurately flag disinformation. Recent strides in autonomous vehicles are encouraging, but technical challenges remain. And, making them cost effective enough for widespread military use will be a significant hurdle. Building a single, robust tactical network to link platforms also remains a much more difficult challenge than it seems. Commercial technologies being adapted for military use might be less technically risky since they are perfected in commercial settings. But, while military performance requirements are often more demanding than commercial ones, the fundamental challenge of being pitted against an actively plotting adversary remains. Brose does not seem to address these issues either.
Technology May Not Even Be the Answer
Finally, we must consider the possibility that the role of commercial technology within the U.S. military may not be desirable in the first place. Brose offers a very specific vision of how artificial intelligence, quantum computing, and networked systems should be wielded by the United States. He paints a detailed picture of sensors that locate adversaries with impunity, a battlefield cluttered with disposable unmanned systems, and networks that will accelerate the tempo of operations to new highs.
This optimistic vision is enthralling, but should it be the goal for which the U.S. military strives? A battlefield network that seamlessly links together sensors and shooters will accelerate the operational tempo — when it works. How will an adaptive adversary seek to disrupt that network and turn its advantage into a liability? How will commanders leverage such connectivity? What role should artificial intelligence play? Will technology enhance initiative and decision-making, further enable micromanagement, or something else?
Brose tangentially examines these issues but only as they concern artificial intelligence and the ethics of armed conflict. He offers a refreshingly nuanced vision of an artificial intelligence that would enhance the abilities of human decision-makers and refrain from making the decisions itself. He forthrightly acknowledges the technological challenges of achieving that ideal. He considers the role of trust and artificial intelligence in military decision-making.
But, Brose never really questions the role of commercial technology and its effect on war in the first place. He admits that the fog of war will never truly lift but still walks readers through a vision of networked warfare where he believes that it does. Some within the defense community urge greater caution about the enthralling vision of networked warfare. Laura Schousboe, B. A. Friedman, and Olivia Garard have argued that the ultimate role of emerging technologies is still unclear. The interaction of humans — both friendly and enemy — and systems should be deliberately considered. Commercial technology is likely to play a significant role in future conflict, but the Pentagon should guard against too much optimism.
No Plan Survives First Contact With the Enemy
Brose has made an important contribution to the debate about commercial technology and the military. He sees the throughline between technologies, their military and political uses, and the domestic organizational and political landscapes. He understands that warfare is an inherently chaotic human endeavor that can defy the expectations of optimistic technologists. As Kill Chain pulls it all together in an admirable way, I hope Brose uses his deep knowledge of defense technology issues to explore the obstacles outlined here as well other ones.
However, technologists and those who share their views should be cautious about how the future of armed conflict will play out. The vision of future war that Brose and others imagine is compelling, but the United States won’t truly know how this situation will play out until a crisis arrives. The same is true for U.S. adversaries. Emerging commercial technologies will play a role, but the military may wish to consider additional steps to make their adoption more effective in the face of such uncertainty.
For instance, the military might consider reforming the requirements process to address the issues of desirability and implementation. Reforming requirements might help the Pentagon fully leverage the flexibility offered by the updated acquisition regulation. This sort of reform can bring clarity to the most useful intersections between emerging technologies and the military, which can also keep cost, schedule, and performance expectations in line with reality.
The Department of Defense can also prepare for inevitable surprises. Richard Danzig observed that predictions about the future of war are consistently wrong. It is better to be circumspect about the nature of future conflicts and prepare for predictive failures. The continued attention to rapid acquisition processes is an encouraging sign. Past experiences with quick responses to unforeseen adversary capabilities also offer lessons to learn.
The radio that Harris Communications built was neither perfect nor the best radio that people could imagine at the time. However, it provided capabilities that were sorely lacking. Its designers accomplished this achievement by combining an understanding of what was technologically possible with a clear grasp of the performance requirements that were most important to users. As the Pentagon and commercial technologists continue to explore the potential of commercial technologies for the military and work towards greater adoption, they may wish to focus not only on lowering bureaucratic barriers but also on managing expectations about what technologies will be most beneficial and how they will be used.
Jonathan Wong is an associate policy researcher at the non-profit, non-partisan RAND Corporation and a non-resident fellow at Marine Corps University’s Krulak Center for Innovation and Creativity. He can be found on Twitter @jonpwong.