Invisible Doomsday Machines: The Challenge of Clandestine Capabilities and Deterrence


Stanley Kubrick’s iconic black comedy Dr. Strangelove remains one of the most insightful works on deterrence. The film revolves around the Doomsday Machine, which will automatically destroy all life on earth if the United States ever launches a nuclear attack on the Soviet Union. After a rogue American general does precisely that, the Soviet ambassador reveals the machine’s existence and explains what is about to happen. American General Buck Turgidson is skeptical, claiming the machine is “an obvious commie trick, Mr. President!” The titular Dr. Strangelove subsequently delivers the film’s biting satirical punch line: “The … whole point of the Doomsday Machine … is lost … if you keep it a secret! Why didn’t you tell the world, eh?”

Hidden within Kubrick’s dark humor is a very real problem: the challenge of clandestine military capabilities. Modern military operations increasingly turn upon elements of military power that depend almost entirely on secrecy for their battlefield effectiveness. But the very secrecy that drives their battlefield impact can interfere with the political objectives military power is meant to serve, like deterrence. The fictional Doomsday Machine was militarily effective — it ensured retaliation after an attack — but it was politically ineffective because it was kept secret. Yet unlike modern clandestine capabilities, the machine’s secrecy was not integral to its military function.

Secrecy can also inhibit the effective integration of foreign policy with military capabilities, detracting from strategic coherence and holding back policy implementation. To succeed in a world of rising military secrecy, U.S. policymakers need to understand the political-military trade-offs posed by clandestine capabilities, develop concepts for when such capabilities should be concealed or revealed, and organize the government internally for managing military secrecy.

The Challenge of Clandestine Capabilities

While military operations generally benefit from secrecy and surprise, most nonetheless retain their utility if revealed. A new tank or missile still functions if an opponent learns about it. Clandestine capabilities, in contrast, may not work at all if revealed to an adversary. Consider alleged U.S. “left of launch” efforts to neutralize North Korean missiles using cyber and electronic warfare techniques. If the North Koreans understand the vulnerabilities these techniques exploit in their systems, they can fix those vulnerabilities — largely or entirely nullifying the left-of-launch capabilities. Similarly, the United States was highly successful in tracking Soviet ballistic missile submarines during the 1960s, but U.S. leaders kept these capabilities secret because of the fear that the Soviets could develop countermeasures (such as making their submarines quieter). When a spy ring revealed this capability to the Soviets, these fears were proved correct. It took years for the United States to recover some of this tracking capability.

The battlefield premium on secrecy makes clandestine capabilities difficult to exploit for political ends such as deterrence, assurance of allies, or shaping a long-term political-military competition. Keeping the capability concealed to protect its military utility makes the capability useless for deterrence — an invisible Doomsday Machine. A middle ground is to simply inform an adversary that you can turn off their missiles or find their submarines while keeping the capability under wraps. But, as Turgidson’s skepticism illustrates, such a claim is likely to be met with incredulity from the adversary. Finally, proving the capability’s existence by demonstrating it may enhance its deterrent effect, but it can also allow an opponent to develop effective countermeasures, eliminating both its military utility and deterrent value. In short, what makes clandestine capabilities militarily useful may reduce their political effectiveness, and vice versa.

The Department of Defense, recognizing this trade-off and the need for deterrence, has decided to reveal some capabilities, as then-Deputy Secretary of Defense Robert Work announced last year: “We will reveal for deterrence, and we will conceal for war-fighting advantage. There are a lot of things in the budget that we don’t talk about because we want to preserve that in case, God forbid, deterrence fails and we do come to a conflict of arms.”

Clandestine capabilities also pose challenges for governments related to the implementation and integration of strategy. One common means to ensure protection of secrets is strict compartmentalization, through what are known as “special access programs,” to which only a limited number of people have access. As an example, most information about the new U.S. heavy bomber, the B-21 Raider, is protected by a special access program, with restrictions on who can be “read in” to the program details. Yet by leaving some policymakers out of the loop, compartmentalization can reduce the effectiveness and coherence of policies that depend on clandestine capabilities. For example, even in classified war games, many participants will not have access to all relevant special access programs. At best, such capabilities appear as what one U.S. wargamer called “magic pixie dust” — participants are told nothing about actual capabilities but are asked to simply trust that they exist and will perform. But commanders and policymakers will be poorly prepared to integrate such capabilities into plans and policies if their exposure (and for senior officials, their staff’s exposure) is limited.

Compartmentalization makes it challenging to integrate and manage clandestine capabilities for deterrence or other political effects. A limited set of individuals may have access to information on the B-21, for instance. These individuals may or may not overlap with those with access to other special access programs, such as those dealing with offensive cyber and space capabilities. Within the Department of Defense there are a number of so-called “super-users” who have access to all the department’s programs, but even these individuals may not necessarily be privy to the intelligence community’s special access programs. The converse is true for many in the intelligence community who cannot access restricted programs from the Department of Defense. Thus the number of individuals with a routine and total picture of U.S. and adversary clandestine capabilities is very, very small. These individuals are typically quite senior and therefore very busy. This presents significant barriers to thinking strategically about integrating an array of clandestine capabilities, ranging from stealth to space to surveillance, and then considering what to reveal and to conceal.

We propose two considerations policymakers could use to navigate the political-military trade-offs inherent in decisions to reveal or conceal clandestine capabilities.

First, policymakers should assess the military uniqueness of a particular clandestine capability and the ability of the target adversary to counter the capability.

At one extreme, concealment for war-fighting advantage makes sense in the case of a unique, irreplaceable capability for which countermeasures are relatively inexpensive. This can require extreme compartmentalization of knowledge of the program. In the early 1980s, according to East German intelligence sources, the United States allegedly developed an electronic warfare program known as CANOPY WING, which was intended to interfere with Soviet command and control systems, including control of strategic nuclear forces. This capability was not signaled to the Soviets because it seemed they could fairly easily fix the vulnerabilities the program exploited (which they may have done with strategic nuclear forces in the mid-1980s). The potential military advantage of this capability in a war, combined with its vulnerability to Soviet countermeasures, made concealment an easy and obvious choice.

At the other extreme, capabilities can readily be revealed when they are not unique and adversary responsiveness is likely to be slow or lackluster. This may have been why the United States decided to reveal its signals intelligence intercepts of Soviet air defense after the downing of Korean Airlines Flight 007 in 1983. The low-level voice intercept capability that was revealed was useful but hardly unique. Moreover, not only would it have been expensive for the Soviet Union to change its communications procedures and encryption, the sclerotic but bureaucratically powerful Soviet Air Defense Forces were likely to resist a major overhaul. The Reagan administration thus chose to reveal the capability to gain the political advantage of exposing the Soviets as lying about how the civilian airliner was downed.

Second, decision-makers should consider options for deception that might avoid the political-military trade-off, at least temporarily. Through effective deception, an actor can reveal a capability’s military effects while concealing the mechanism by which it functions, potentially bolstering deterrence while preserving military utility.

The United States apparently did this successfully in the 1980s, when it revealed the efficacy of its stealth technology to the Soviets while deceiving them about the exact nature of stealth. After French intelligence obtained the “Farewell Dossier,” which revealed the bulk of KGB intelligence collection against Western technology, the United States was allegedly able to use the information to tailor a counterintelligence and deception campaign to feed the Soviets false information. In 1986, possibly thanks to intelligence revealed in the dossier, a U.S. Air Force officer helped snare a Soviet air attaché seeking information on stealth. Armed with an understanding of what the Soviets were looking for, the United States sought to mislead them. One of the chief architects of this deception campaign for the CIA claims, citing unclassified sources, that the “Pentagon introduced misleading information pertinent to stealth aircraft, space defense, and tactical aircraft.”

This misleading information likely contained enough detail on actual capabilities to let the Soviets know stealth was a real and growing U.S. capability without letting them understand enough to negate it. A CIA assessment in 1988 concluded “Soviet knowledge of U.S. Stealth systems … has allowed the Soviets to better anticipate what offensive threats they will face in the future and possibly to focus research on counter low observable (CLO) systems. We have no evidence of a Soviet CLO system, however.” Thus, at least through 1988, partial revelation had not led to an effective countermeasure to stealth, yet may have bolstered deterrence.

Organizing and Managing Clandestine Capabilities

As the United States increasingly contemplates acquiring clandestine capabilities for deterrence and warfighting, it should consider rethinking its approach to managing these programs to better avoid trade-offs. While the Department of Defense has made progress in making some special access programs more widely accessible, integration of these so-called “black” capabilities across the government still seems to be a challenge.

As a first step, policymakers should consider unifying and refining the oversight and management of such programs, which are currently divided between committees in the intelligence community and the Department of Defense. While membership of these committees overlaps in some cases, a handful of people with adequate knowledge of clandestine capabilities is not sufficient to integrate policy across the whole of government. A unified committee would need a modest but non-trivial number of full-time staff with broad access to special access programs, since the senior leadership of the committees — including the Deputy Secretary of Defense, Vice Chairman of the Joint Chiefs of Staff, and the Principal Deputy Director of National Intelligence — are far too busy to manage oversight on their own.

While it may be bureaucratically uncomfortable, a unified approach to management would enable government organizations to create a common framework for considering the trade-offs of concealment versus revelation. This unified approach could also improve the ability to integrate clandestine capabilities within government, potentially improving evaluations of capability uniqueness and enemy responsiveness as the government decides whether to reveal or conceal. Moreover, it would help the government consider the possibility of partial revelation through counterintelligence and deception. In short, a unified management approach would go a considerable distance toward meeting the challenges of clandestine capabilities. In a world where emerging technologies are enhancing the salience of “black” programs relative to “white” ones, the U.S. government must adapt its methods.


Brendan Rittenhouse Green is an assistant professor of political science at the University of Cincinnati. Austin Long is a senior political scientist at the nonprofit, nonpartisan RAND Corporation. Their research on this topic was funded by the Carnegie Corporation of New York and the Defense Threat Reduction Agency’s Program on Advanced Systems and Capabilities for Combating Weapons of Mass Destruction.

Image: U.S. Air Force/Alyssa C. Gibson