The Kill Chain

Christian Brose spent the last years of Senator John McCain’s life as the staff director of the Senate Armed Services Committee, watching from the inside as the American defense establishment approved budget after budget for the same kinds of ships, tanks and fighters it had bought for decades. The Kill Chain, published by Hachette Books in the spring of 2020, is his account of why he believes that pattern will end badly. Brose writes as a former political appointee — foreign policy speechwriter to Condoleezza Rice, then McCain’s foreign policy adviser and finally his committee chief — who had already moved into industry as chief strategy officer of the defence-technology firm Anduril by the time the book appeared. The tone throughout is that of a lifelong Republican national security hand who has decided that the institution he served is no longer telling itself the truth. It joined a conversation that had been building since the mid-2010s, when Robert Work’s Third Offset strategy first put artificial intelligence and human-machine teaming at the centre of Pentagon thinking, and when Elbridge Colby and the authors of the 2018 National Defense Strategy pivoted formally from counter-terrorism to great-power competition with China and Russia.

The argument at the heart of the book is that the American military has become extraordinarily good at building a small number of extraordinarily expensive platforms — the F-35 fighter, the Ford-class aircraft carrier, the Zumwalt destroyer, the Ground Combat Vehicle — while the People’s Liberation Army has spent twenty years designing the tools to hunt exactly those things. Brose describes the resulting force as the one the country needed for the last two decades of counter-insurgency, built for uncontested skies over Iraq and Afghanistan, and warns that it is now brittle where it needs to be resilient. The lens he uses is the kill chain itself: the loop of understanding what is happening, deciding what to do, and acting on it. Modern militaries, he argues, win when they close that loop faster than the other side. American practice has been to build ever more capable individual nodes — the sensor that sees further, the missile that flies faster — while neglecting the networked flow between them. China’s answer has been the opposite: dense fields of sensors, missiles and command systems designed to fracture the American kill chain by killing the tankers, satellites and command aircraft it depends on.

The book is structured as a diagnosis moving into a prescription. The opening chapters set the historical stage, sketching how the American way of war took its current shape after the 1991 Gulf War, when precision-strike weapons and networked sensors produced the Revolution in Military Affairs that other countries then studied and copied. Brose spends considerable time on Andrew Marshall’s Office of Net Assessment, the small Pentagon shop that spent decades warning about exactly this dynamic, and on the Chinese military writers — most notably those behind the “assassin’s mace” concept and the anti-access, area-denial doctrine that shapes the People’s Liberation Army Rocket Force. Middle chapters walk through the specific American procurement failures Brose watched from the committee: the Army’s Future Combat Systems, cancelled after tens of billions of dollars; the Littoral Combat Ship, whose two competing hull designs neither found a mission nor survived contact with the fleet; the long, painful gestation of the F-35 program; the Ford-class carrier’s troubled electromagnetic launch and arresting gear. Later chapters turn to what he thinks the answer looks like: smaller, cheaper, more numerous unmanned systems; open architectures instead of proprietary ones; software that can be updated the way a smartphone app is updated rather than a decade-long block upgrade; and, most controversially, a much larger role for machine autonomy inside the kill chain itself. The book closes with chapters on the political economy of defence — how the congressional appropriations process, the concentration of the prime contractor base, and the culture of the requirements process combine to keep producing the wrong things — and on what Brose calls a moving target for the next generation of American strategists.

The concrete evidence Brose brings is what carries the argument, and it is the part of the book that has aged best. He describes the way modern anti-ship ballistic missiles — the Chinese DF-21D and DF-26, often labelled “carrier killers” — invert the maths of a carrier strike group by making a ship worth thirteen billion dollars vulnerable to a weapon costing a small fraction of one percent of that. He walks through the vulnerability of the tanker fleet on which every long-range American strike depends, the KC-46 and KC-135, most of them unstealthy and orbiting predictable tracks. He discusses satellites and the growing Chinese and Russian counter-space arsenals — direct-ascent anti-satellite weapons of the kind China first demonstrated in 2007, co-orbital inspectors, and the ground-based lasers the Pentagon has repeatedly warned about. He examines the E-3 Sentry AWACS and the Joint STARS ground-surveillance aircraft, both of which he argues are dead men flying in any real fight with a modern power. On the other side of the ledger he draws examples from the small world of Silicon Valley defence firms and from DARPA — programmes like the Sea Hunter unmanned surface vessel, the loyal-wingman concepts that would later surface in the Air Force’s Collaborative Combat Aircraft effort, and the CODE program for co-operating autonomous swarms. He tells the story of the failed Project Maven rollout inside Google as a parable of the cultural gap between the technology industry and the Department of Defense, and points to Palantir’s legal battle with the Army over the Distributed Common Ground System as a case study in how the acquisition system fights software firms.

The book also draws on people. John McCain is a presence throughout, both as a mentor whose loss the author feels and as an example of the political will Brose believes the reform effort now lacks. Andrew Marshall appears as a kind of Cassandra figure, retired by the time of writing and dead not long after. William Perry, Robert Work and Ashton Carter — three former deputy or full defence secretaries with sharply differing styles — are treated as the last generation of civilian leaders to seriously try to steer the Pentagon toward new technology, with mixed results. Brose is generous to Silicon Valley figures like Palmer Luckey and Trae Stephens who have moved into defence, and cautious about the industry giants — Lockheed Martin, Boeing, Raytheon, Northrop Grumman, General Dynamics — whom he describes as trapped by the same incentives that trap their government customer. He is unusually candid about congressional dynamics, having watched them from the inside, and singles out earmark politics, service parochialism, and what he calls the military-industrial-congressional complex as reasons the wrong programmes persist.

Reception among defence writers and practitioners has been broadly favourable but not uncritical. The book landed in the same season as several works making adjacent arguments — Michèle Flournoy and Michael Brown’s writings on defence innovation, T. X. Hammes on cheap distributed systems, Paul Scharre’s Army of None on autonomy in weapons — and it has been read alongside them as part of a coherent reform school of thought. Its influence on the Pentagon’s own Joint All-Domain Command and Control initiative is visible: JADC2, formally endorsed by the Joint Chiefs in 2021, tries to do exactly the kind of networked, cross-domain sensor-to-shooter fusion Brose argues for, though how much of that owes directly to the book and how much to parallel thinking inside the department is hard to say. Critics have pushed back on several fronts. Some have argued that Brose overstates the novelty of his kill-chain framing — that the language of observation, decision and action is essentially the Boyd loop of the 1970s dressed in modern clothes. Others have argued that his enthusiasm for small, cheap, autonomous systems underestimates the enduring value of exquisite platforms in specific missions, particularly undersea, and that a fleet of drones cannot substitute for the deterrent presence of a carrier. There is also a class of criticism that reads the book as a business case for a segment of the defence industry the author now works in, and that the reform prescription is easier to write from Palo Alto than to enact through the House Armed Services Committee. Brose is open about the transition; readers should judge the book knowing where its author sits.

For a reader working through the field of AI in warfare today, The Kill Chain sits in a specific and useful spot. It is not the book to read on the ethics of autonomous weapons — Scharre and, more recently, Kenneth Payne cover that ground more directly — and it is not the technical book on how modern military artificial intelligence actually works. What it is is the political and institutional book: a plain-language account of why a rich country with the world’s most powerful military can still find itself outbuilt in the systems that matter, and what changing that would take. It pairs well with Scharre for the ethics, with Hammes for the tactical logic of cheap distributed systems, with the RAND studies on Taiwan wargaming for the operational stakes, and with anything on the Chinese military writings of the past two decades — the Science of Military Strategy in particular — for the other side’s view. It does not cover the drone war in Ukraine, which came after publication and which has vindicated some of its predictions while complicating others; nor does it engage seriously with the European defence base, which Brose treats as a footnote to the American picture.

The durability of the book will depend on what happens next in the Pacific. If the current American shift toward distributed maritime operations, loyal wingmen, autonomous underwater vehicles and the Replicator initiative announced by Deputy Secretary Kathleen Hicks in 2023 continues and produces results at scale, The Kill Chain will read as an early and clear statement of an idea that eventually won. If the Pentagon proves unable to break the procurement habits Brose describes, and a real crisis exposes the vulnerabilities he catalogues, the book will read as a warning that was on the shelf when it might have been useful. Either way its diagnosis — that the American military is optimised for a war it has already fought and unprepared for the one it might have to — is likely to outlast most of the specific programmatic prescriptions in it, and the kill chain as a way of talking about competition between modern forces has entered the vocabulary of the field.