TY - JOUR
AU - Schulzke,, Marcus
AB - Abstract The controversy surrounding military drones has generated many proposals for restricting or prohibiting existing drones, additional autonomous variants that may be created in the future, and the sale of drones to certain markets. Moreover, there is broad interest in regulating military drones, with proposals coming not only from academics but also from NGOs and policymakers. I argue that these proposals generally fail to consider the dual-use character of drones and that they therefore provide inadequate regulatory guidance. Drones are not confined to the military but rather spread across international and domestic security roles, humanitarian relief efforts, and dozens of civilian applications. Drones, their component technologies, the control infrastructure, and the relevant technical expertise would continue to develop under a military-focused regulatory regime as civilian technologies that have the potential to be militarized. I evaluate the prospects of drone regulation with the help of research on other dual-use technologies, while also showing what the study of drones can contribute to that literature. Drones’ ubiquity in nonmilitary roles presents special regulatory challenges beyond those associated with WMDs and missiles, which indicates that strict regulatory controls or international governance frameworks are unlikely to succeed. With this in mind, I further argue that future research should acknowledge that drone proliferation across military and civilian spheres is unavoidable and shift focus to considering how drone warfare may be moderated by countermeasures and institutional pressures. Drones, security, weapons Introduction The dual-use problem, which arises when the same commodities can be used in weapons or in peaceful civilian applications, is a challenge for research on international security and international trade that brings these fields into dialogue with each other and forces us to make tradeoffs between economic, humanitarian, and security goals (Fuhrmann 2008). “Most of the components of major weapons systems are dual use in nature” (Fuhrmann 2012, xi), making it difficult to prevent the spread of weapons technologies without also impinging on legitimate trade or even interfering with the acquisition of goods that have vital peacetime benefits. Further complicating the dual-use problem is that the knowledge and material infrastructure acquired from civilian programs can provide the foundation for later efforts to build weapons. Most of the research on dual-use technologies focuses on weapons of mass destruction (WMDs) and ballistic missiles (Mistry 2003; Busch and Joyner 2009; Stinnett et al. 2011; Fuhrmann 2012; Tucker 2012b), but the proliferation of drones over the past decade presents a new regulatory puzzle with distinct challenges that must be considered to advance the study of dual-use technologies and, in particular, research on drones. Drones have the potential to alter how and when wars are fought, especially when it comes to asymmetric conflicts (Horowitz, Kreps, and Fuhrmann 2016). By allowing their users to strike from a distance and avoid the risks associated with deploying human combatants on the battlefield, drones arguably lower the threshold for using lethal force and make it easier to escape accountability for use-of-force decisions (Kreps and Kaag 2012; Sauer and Schörnig 2012; Enemark 2013; Kaag and Kreps 2014; Chamayou 2015). Drone variants operating in other roles or those with higher degrees of autonomy raise still more worries, including the prospect that autonomous machines may one day be able to kill independently (Sparrow 2007; Krishnan 2009; Tonkens 2013; Schulzke 2013). Concerned academics, policymakers, and activists have proposed a range of regulatory mechanisms to curtail the use of UAVs and the development of new drone models (Altman 2013; Zenko and Kreps 2014; Boyle 2013; Buchanan and Keohane 2015; Sparrow 2016). However, these efforts suffer from a serious defect: they assume a sharp division between military and civilian technologies. Research on the dual-use problem must address a larger range of innovations, take a broader conception of dual-use that includes software and finished products in addition to the component parts that are typically the object of study, and recognize the importance of cultural and normative shifts that may alter attitudes toward dual-use goods. At the same time, the study of drone regulation should be situated within the research on dual-use technologies and weapons regulation. Research on military drones is usually limited to machines that are designed with explicit military functions, and even here the focus is on machines that are armed, rather than on the many types of unarmed military drones. Conversely, research on drones and autonomous systems in civilian roles, such as in cars, proceeds without much comment on how these could impact the future of international security (Meier 2015). Looking at the dual-use character of drones exposes the unrecognized challenges inherent in existing regulatory proposals, while also offering opportunities for rethinking how future research could be more profitably framed. Drones pose several regulatory challenges that go beyond those coming from the dual-use goods associated with WMDs and ballistic missiles or even those related to more conventional military hardware. First, drones are dual-use in a more expansive sense than how this term is generally used. Most efforts to restrict dual-use technologies focus on the component parts of weapons, while treating final products as being military or civilian in their function. Drones are dual-use, both when it comes to their constitutive parts and when it comes to the final products, because machines produced for nonmilitary security roles or the commercial market can be transformed into weapons or vice versa, often with little modification. Second, the market for drones is so vast, heterogeneous, and conducive to civil-military technology sharing that it is open to a more diverse array of actors than other dual-use markets. This increases the possible routes of technology sharing, thereby raising additional regulatory challenges. The diversity of drone models also introduces formidable difficulties when it comes to defining the scope of regulations. Broadening restrictions in response to this challenge would improve the likelihood of successfully curtailing the proliferation of military drones but would come at the expense of increasing the costs of restrictions and the incentives for defection. Third, the pervasiveness of drones and related technologies in civilian life, as well as their usefulness in humanitarian roles, could cause cultural shifts beyond those associated with dual-use goods that are less visible in civilian life. Finally, one of the central problems for proponents of drone restrictions is the effort to prevent the development of autonomous drones that could select and engage targets by themselves or with limited human oversight. Regulatory efforts might therefore have to focus on software in addition to hardware. I argue that ambitious projects to severely constrain or even prevent military drone proliferation are untenable because they cannot cope with the dual-use problem. Future research would benefit from exploring more modest efforts to slow the pace of technological diffusion. Although drones are potentially revolutionary weapons platforms, the impact of current generation drones has been limited because they are fairly unsophisticated compared to manned vehicles and because they are vulnerable to countermeasures (Horowitz, Kreps, and Fuhrmann 2016). This suggests that drones present a security risk that can be managed so long as the machines do not advance too quickly. Restrictions that are designed to moderate the pace of development, rather than stopping it entirely or creating elaborate oversight mechanisms that states would have little incentive to support, could make it possible to realize the benefits of drones while still affording the time needed to develop effective countermeasures for those that are directed into military purposes. Such restrained diffusion would not prevent drone proliferation, but it would have the advantage of being an easier point of compromise between the many private and public actors that would need to coordinate to limit the industry. Subsequent research on drone regulation would be best served by looking for ways of managing drone proliferation across military and civilian roles so that it does not dramatically upset the balance of military forces. It is particularly important to give more thought to the role of antidrone countermeasures and to institutional pressures that may slow the adoption of drones in military roles. The Dual-Use Problem The dual-use problem is endemic to military technologies. Most weapons and military vehicles depend on components, technical expertise, and/or production capacities that can also be used in civilian goods, including those that are vital to energy infrastructure, agriculture, and medical treatment. Even well-intentioned investments in dual-use goods for civilian purposes can become a starting place for weapons development if those intentions change. Fuhrmann (2012) finds that India, Pakistan, and North Korea were each able to develop nuclear weapons by drawing on foreign aid directed at peaceful nuclear programs and argues that investments in peacetime industries lay the foundation for more rapid acquisition of arms. Similarly, Tucker argues that as dual-use technologies linking chemical and biological warfare to agriculture and medicine become more commercialized, the requisite knowledge needed to operate them will decline and “they will become more accessible to terrorists and other malicious actors who possess only basic scientific skills” (Tucker 2012a, 24). Studies of dual-use technologies have identified several major challenges when it comes to restricting weapons proliferation and preventing civilian goods from being misused. Many goods require end-use agreements that specify how they will be employed, but these can be faked and are poorly monitored (Corera 2006, 118). States may lack the capacity to police dual-use goods, as was the case in former Soviet states during the 1990s (Bertsch and Grillot 1998). They may likewise lack incentives to support regulatory regimes. “Legitimate trade in dual-use goods can be quite profitable, making it potentially costly for governments to restrict such trade” (Stinnett et al. 2011, 301). Unlike conventional military hardware, it is also possible to conceal the transfer of dual-use goods by either genuinely intending them to benefit civilian programs or giving the false appearance of this intent. Even with regulatory regimes in place, states have strong incentives to free ride because they may benefit from controls enacted by others (Stinnett et al. 2011, 312). States that formally support restricting dual-use goods may fail to make meaningful contributions to that end when this would incur costs on themselves. There are also strategic benefits associated with defection. States may gain a decisive advantage by pursuing new weapons technologies when rivals are blocked from doing so. They may likewise assist allies, enemies of enemies, and states with similar regime types by providing peaceful dual-use assistance that can be channeled into military purposes. These interests are so powerful that Fuhrmann (2009, 183) finds that “supplier states’ strategic interests are more important than limiting the spread of nuclear weapons.” Despite making valuable contributions to the fields of international security and international trade, studies of dual-use goods focus disproportionately on WMDs and ballistic missiles. In an era when military innovation is often in pursuit of “useable weapons” (Koplow 2010) that inflict less destruction and that are legitimized with appeals to increased precision, it is vital to broaden the scope of research. Drones offer a particularly important area of study because they raise an array of moral, legal, and policy concerns related to an overreliance on precision weapons—a much different challenge than the one presented by immensely destructive WMDs (Kreps and Kaag 2012; Sauer and Schörnig 2012; Enemark 2013; Kaag and Kreps 2014; Chamayou 2015; Schulzke 2017). Moreover, as I will show, drones raise special regulatory problems that go beyond those identified in previous work on dual-use commodities. Proposals for Regulating Drones I use the term “drone” in a broad sense to refer to vehicles that can operate without an onboard pilot. Most current generation drones are remotely piloted, permitting a human controller to direct the machine without being physically present. The most famous drones in military and civilian domains are unmanned aerial vehicles (UAVs) or remotely piloted aircraft (RPAs), but some variants can also operate on land and at sea. I use the term “autonomous drone” when talking specifically about drones that can operate without direct guidance from a human operator. Drones of various types can be grouped together as a large family of related technologies that make use of many of the same software and hardware components. Above all, they are defined by the capacity to operate as proxies for humans and piloted vehicles, ensuring that the controller is safe regardless of what happens to the drone. This capacity for displacing risks and minimizing human involvement is one of the reasons why drones are so attractive in conflicts. The ability to avoid committing soldiers can save lives and prevent embarrassing international incidents involving captured pilots. Military UAVs and the targeted killing missions that have come to define their contribution to the War on Terror are justifiably controversial. Critics have raised dozens of arguments against using these machines, from concerns that they will make it too easy to wage wars to claims about how they may erode normative constraints on fighting (Enemark 2011, 2013; Sauer and Schörnig 2012; Kaag and Kreps 2014). These objections range from hypothetical scenarios involving future conflicts to objections against how the United States, United Kingdom, and Israel have used UAVs in targeted strikes against suspected terrorists. They look back on recent operations involving mistaken attacks on civilians and forward to hypothetical futures in which drones dominate the battlefield in ways that confer unfair advantages on technologically advanced states. Moreover, opposition to the spread of drones comes not only from academics but also NGOs, policymakers, journalists, and concerned citizens (Kilcullen and Exum 2009; Benjamin 2012; Woods 2015; Shane 2015). There is therefore broad concern about military drones that brings many different types of actors together in a shared effort to minimize the impact of these machines that threaten to irrevocably change the nature of war (Singer 2009, 2015). Commentators defending drones attempt to show that they are justifiable tools for limiting violence against civilians, protecting soldiers, and more accurately striking terrorists (Arkin 2009; Strawser 2010; Schulzke 2011, 2014; Byman 2013). However, detractors insist that there are strong prudential grounds for supporting regulation. Even though many of the antidrone arguments are speculative and rest on predictions about what might happen in future conflicts, proponents of regulation contend that the possibility of drones destabilizing international security warrants caution. With this in mind, they have introduced dozens of proposals to curtail drone use, promote accountability, discourage proliferation, and prevent further innovations. Buchanan and Keohane (2015, 17) outline an international drone regulatory regime that is “designed to enhance accountability through transparency and publicity and, in countries where this is feasible, to mobilize domestic constituencies to support compliance with the regime.” They acknowledge that states have good reasons for employing drones but maintain that it is essential to develop some degree of international accountability to discourage states from violating each other's sovereignty, overusing drones, or evading responsibility for the consequences of attacks. According to their Drone Accountability Regime, “each state would establish a national supervisory body with explicit, public procedures for ensuring that the template for regulatory supervision accepted by the state is actually implemented” (Buchanan and Keohane 2015, 29). Recognizing that states may sometimes feel compelled to launch quick attacks without waiting for approval, Buchanan and Keohane go on to outline provisions for obtaining post facto permission for strikes. This proposal is primarily aimed at limiting targeted killings, rather than with preventing the use of drones in general. However, Buchanan and Keohane do mention a more general need for inhibiting the spread of drone technologies to ensure that these weapons platforms remain under state control. They contend that the Drone Accountability Regime could have the power to prevent violent nonstate actors from developing armed drones. Here we see an effort to circumscribe the range of actors that have access to drones in an attempt to reinforce the Westphalian system of interstate politics that their proposal depends on. Boyle (2013) takes a more skeptical view of drones, saying that many of their apparent advantages are overstated and that they introduce serious risks that interfere with responsible use. He is also concerned that the proliferation of drones will ultimately lead the United States to undermine its own interests. Although it has benefited from a near-monopoly on drones in recent conflicts, it will find that this short-term gain comes at the expense of “the traditional concepts which have underwritten stability in the international system” and emerging threats from rivals armed with their own drones (Boyle 2013, 23). He admits that it would be unrealistic to think that drones can be prohibited now that they play such a prominent role in conflicts but argues that serious attention must be given to enacting regulations. Boyle (2013, 28) is especially interested in creating a UN investigatory body that could “ensure that their use and sale are transparent, regulated and consistent with internationally recognized human rights standards.” He points out that the United States has an incentive to facilitate restrictions while it can still act from a position of strength as the world's foremost drone user. Zenko and Kreps (2014) contend that the spread of military drones could allow states to target domestic opponents or that their use could make it easier to escalate conflicts. They argue that antiproliferation measures are essential for preventing a future of increased domestic and international conflict. This concern over domestic spaces, in addition to the international, sets them apart from other commentators who are exclusively interested in the latter. Zenko and Kreps also argue that there are some natural constraints on drone proliferation. First, the technologies needed to produce advanced drones are expensive and require extensive support structures to employ effectively (Zenko and Kreps 2014, 14). Second, deploying drones internationally requires foreign airfields or aircraft carriers (Zenko and Kreps 2014, 15). Third, domestic opposition can interfere with the creation and deployment of new weapons. Zenko and Kreps echo Boyle in arguing that the United States should take the lead in enacting regulations. They therefore extend the debate over drones to cover international and domestic uses while also offering some hope that regulation may be a realistic goal because of structural limits. Much more could be said about the many different regulatory proposals that have been advanced for limiting the potential costs of using drones. What is important from these examples is the volume and diversity of these proposals. They are directed at existing and future drones, at operations by states and nonstate actors alike, and with attention to international and domestic security contexts. They tend to assume that problems linked to drones could be managed if only there were broader state-level agreement that regulations should be enacted. Additionally, most of these arguments suggest that the United States is the linchpin in enacting new restrictions, such that securing its support might be enough to bring order to drone warfare around the globe. As I show over the following sections, we should be less confident about the prospects of regulation and of the ability individual states have to contain drone use because of the special problems arising from drones’ overlapping civilian and military applications, and future research on drones needs to take this skepticism into account. Drones in Nonmilitary Security Operations The dual-use character of drones, both in terms of drones’ constitutive parts and in terms of the finished machines, exposes a serious weakness in the kinds of regulatory proposals that have been developed. Regulations that focus exclusively on military drones or on certain military applications of drones, such as in targeted killings, seem to assume that drones naturally belong to the domain of military technology and that they are largely insulated from nonmilitary arenas. Over the following sections I show that this assumption is flawed and that drone use for civilian and military roles are already so entangled that it is difficult to imagine any purely military approach to regulating drones succeeding. Drones are pervasive outside of the military, being used for dozens of security, commercial, and entertainment purposes. Each of these domains reveals the extent to which drones and their associated technologies are advancing without falling under the scope of regulations designed specifically for military hardware. Looking at the pervasiveness of drones in nonmilitary settings highlights the importance of moving beyond the existing framework for studying dual-use technologies, which tends to focus on the constitutive parts of weapons, rather than on finished products. Attention to components makes sense when there are vast differences between the goods’ military and civilian applications, which is usually the case when it comes to the WMDs that feature most prominently in the regulation literature. However, the transferability of completed drones across civilian and military functions with comparatively few modifications poses special regulatory challenges. It means that drone hardware and operational knowledge can develop in nonmilitary settings and that technologies can be transferred between military and civilian domains as expediency requires. The use of drones in nonmilitary security settings is particularly important, as the machines involved are often similar, or even identical, to those used by militaries and perform comparable roles. Perhaps the clearest example of civil-military exchange in drone operations comes from the use of UAVs for border security and domestic law enforcement. The American Border Patrol (ABP), a nongovernmental organization that attempts to intercept migrants crossing the US-Mexico border, deployed a small surveillance drone in April 2003 (Bunn 2003). The group used its drone to demonstrate the feasibility of UAV border patrolling and to pressure the US government to adopt this approach. By November of that year Immigration and Customs Enforcement (ICE) was testing unarmed ex-military Reaper drones along the border. Since 2004, Customs and Border Protection (CBP) has conducted patrols using Reaper drones, having nine in service by 2016 and formulating plans to procure more (Nixon 2016). From this we can see that drones manufactured for civilian uses (as the ABP's aircraft was) and for military uses (as the Reaper was) can be used for the same kinds of operations. We can also see how easily military drones were shifted into a civilian role by simply having their weapons removed, evidence that the machines are readily adaptable to an array of operations. If drones can be transferred into civilian roles by having weapons removed, then machines developed primarily for nonmilitary security roles that fall outside the remit of regulatory proposals could likewise be militarized with the addition of weapons systems developed for conventional aircraft. Drones are also being used for border security beyond the United States. The United Kingdom has used drones operated by the Royal Navy to assist in its searches for migrants crossing the Mediterranean (Farmer 2015). These are deployed on an ad hoc basis and remain under military control, yet they show the ease with which these machines can transition between armed and unarmed security roles. The European border security agency Frontex plans to employ drones on a more permanent basis and has been seeking permission to do so from the European Union for several years (Csernatoni 2016). This demand, combined with the existing drone surveillance from the British military and civil society organizations indicates that Europe will soon be using UAVs to secure its borders and that it will have a domestic production infrastructure that could be directed into military purposes. The use of drones for domestic security has not stopped at borders. The CBP's UAVs are regularly loaned out to other law enforcement agencies—from thirty times in 2010 to 250 times in 2012—though this borrowing is now becoming unnecessary as police purchase their own unmanned vehicles. Dozens of police departments across the United States now have permanent UAV fleets, often with no clear restrictions on how they can be employed (Waddell 2015). In 2015, North Dakota became the first state to legalize drones armed with nonlethal weapons (della Cava 2015). Once again, this trend is evident cross-nationally. The United Kingdom has been testing police drones since 2007, with thoughts about using them in dangerous areas and for crowd control (BBC 2007). More recent plans call for the expansion of drone forces across the UK and for broadening their operations to include forensic investigations and patrolling (Wired 2016). Other countries experimenting with police drones include Japan, India, Germany, China, and Canada. Most of these borrow from military designs and may in turn inform the creation of military drones in the future, given how closely military drones are already linked to those used for domestic security. This kind of sharing is apt to continue, or even increase, if military drone development is halted or severely limited by new regulations. If law enforcement were to become the main forum for exploring the possibilities inherent in unmanned systems, then a predictable outcome would be more investment in domestic security drones that have the potential to be militarized by states defecting from regulations. Law enforcement drones are already outpacing their military counterparts in certain respects. For example, China currently appears to be leading the way in ground-based policing with the Anbot, a machine that is roughly the size of a person and armed with nonlethal weapons. If deployed, it would be able to navigate and respond to threats without human control, making it more independent than existing military drones (Lin and Singer 2016). Police drones could be especially attractive to armed forces experimenting with nonlethal counterinsurgency techniques that are already heavily informed by law enforcement technologies and practices (Koplow 2006). Such machines could potentially be deployed in counterinsurgency contexts without needing to be armed. Drones can do valuable work promoting human security, even saving lives under conditions that might be too dangerous for rescue vehicles with onboard crews. Humanitarian organizations have used UAVs to detect wild fires, deliver supplies to people in remote areas, search for migrants crossing rough seas, and track the path of natural disasters (Meier 2015). These roles are especially important for NGOs, allowing them to operate even when they are impeded by personnel or funding shortages. It is imperative for this kind of work to continue, yet it creates a dilemma. Broad restrictions on drones that cover military and civilian domains could cost lives by hindering humanitarian ventures, but restrictions that focus narrowly on military drones would create the risk of humanitarian drones being armed or of their constituent parts being appropriated for military purposes. Any machine that can more effectively locate refugees or carry food to them could be adapted to locate suspected terrorists or to carry bombs. Arming “Nonlethal” Drones The use of drones for border security, law enforcement, and humanitarian operations shows that similar machines and their constitutive technologies are not relegated to military operations. Moreover, the same basic operations, such as tracking, patrolling, and delivering payloads, are undertaken across civilian and military roles. Curtailing the development of military drones without addressing drones in these other domains would not succeed in halting the pace of innovation. Such regulations would leave open the possibility of the machines being developed for border security, law enforcement, and even humanitarian purposes growing in sophistication and providing a readymade source from which any military defecting from a governance regime could borrow. Thus, even if regulations focused on military drone operations were successful, the availability of drones in other sectors would provide the technologies and skills that would be needed for quick remilitarization. We can see from past drone innovations that this kind of borrowing is a genuine concern. The drone was originally designed as an unarmed surveillance aircraft and was only outfitted with Hellfire missiles nearly a decade after its first deployment (Whittle 2014). American police officers were able to quickly convert a bomb disposal drone into a bomb delivery platform during a standoff in 2016 (Edwards 2016). This kind of improvisation makes virtually any unarmed drone a potential weapons platform. Reading through the literature on drones one gets a sense of the confusion they are already creating by obscuring the boundaries between different security contexts. Dunn contends that “[t]hey blur the distinctions between the military and intelligence worlds, between warfare and law enforcement, between combat and assassination; between the battlefield and the hinterland, between the territories of allies and enemies, between domestic and foreign threats, and between counterterrorism and counter-insurgency” (Dunn 2013, 1238). As he correctly notes, drones have been used for an array of security roles that draw military and intelligence services together. They epitomize the close, and often secret, interagency relationships that have developed throughout the War on Terror as the United States and its allies have sought new ways of striking first against their adversaries. Already the confusion between war and policing interferes with the “war paradigm” versus “policing paradigm” distinction that is central to Buchanan and Keohane's (2015) drone regulatory regime. We must go one step further by recognizing that this confusion of paradigms is not something that can be easily cleared up by reemphasizing the conceptual distinction between the two. It is a confusion embedded in practices of dual-use technology sharing that are becoming entrenched. Drones are not alone in being amenable to use in military and nonmilitary security roles. Other aircraft, ships, and ground vehicles may likewise have multiple applications. What sets drones apart from other technologies is that they have been singled out by so many commentators as being uniquely dangerous pieces of military hardware that should be subject to special restrictions. The spread of finished drones and of their constitutive parts into nonmilitary security applications is evidence that extensive borrowing is taking place. Any efforts to restrict drone innovation in one domain of security operations would be incomplete without looking at the production and use of complementary or even identical technologies in other roles. However, imposing restrictions on those other roles, and especially in humanitarian operations, would impose high opportunity costs and could indirectly cause immense suffering. This would also raise the challenges of gaining ascent to regulatory programs and the potential rewards for defecting from any agreements that are reached. The Many Faces of Civilian Drones Drones designed for the civilian market are already being deployed in conflicts, either as components in weapons systems or as finished vehicles. Control systems for improvised explosive devices (IEDs) are sometimes constructed using those taken from remote controlled vehicles (Clutterbuck 1994, 51–52). Hezbollah has experimented with drones converted from civilian remote-controlled aircraft, showing that fairly primitive UAVs can have offensive applications (Hoenig 2014). Al Qaeda affiliate Jund al Aqsa has used small drones to drop bombs on Syrian soldiers—even posting videos from the aircraft online (Joscelyn 2016)—and Islamic State has resorted to using drone aircraft and cars to conduct reconnaissance and deliver explosives (Hambling 2016). Direct appropriation of civilian drones is attractive for non-state actors that lack the resources to produce their own machines, but state military forces also draw assistance from the civilian market. The Iraqi Army has emulated Islamic State by using drones to guide artillery fire and to conduct reconnaissance (Bodeti 2016). Even the US military regularly incorporates interfaces taken from videogames and civilian drones because this is more cost effective and user friendly than designing new interfaces (Singer 2009, 68; Stahl 2009, 91). With these precedents in mind, it is plausible that armed forces and violent nonstate actors could borrow machines from civilian industries in the future and convert them into weapons platforms. And the degree of borrowing could easily increase if any regulations put armed forces under production constraints. A possible response to the challenge of regulating drones is to extend the coverage of restrictions beyond the military to include machines in nonmilitary security and civilian applications. However, even if it were possible to identify and ban all civilian drones that could have military functions, such a strategy would raise more problems than it would solve. For one thing, casting the regulatory net broadly would increase the challenge of gaining ascent. States may be willing to forgo having armed military drones, but abandoning remote technologies altogether is a far more onerous request. This could put any restrictive states at an economic disadvantage by prompting drone manufacturers and companies wishing to use drones to shift their operations elsewhere. Civilian drone manufacturers have already shown that they are adept at escaping regulations by relocating production. Amazon first explored the idea of Amazon Prime Air, a delivery service using autonomous drones to reach customers in as little as thirty minutes, in 2013. It came up against US legislation prohibiting the use of UAVs by commercial entities, but rather than stopping Amazon, this only compelled it to search for new markets abroad (Vanian 2016). Dominos has followed a similar course, testing its autonomous aerial and ground based pizza delivery drones outside the United States (Bolton 2016). Now there are fears that if the United States does not reevaluate its stance on commercial drones, companies could shift their research and development programs to countries that are more amenable to them (West 2015). The mobility of drone producers generates powerful incentives to abandon domestic regulations in an effort to stay competitive. West (2015) cites estimates showing that the commercial drone market will be worth around $1.7 billion by 2025. Any state that restricts the civilian market would increase the possible gains for defectors that remain open for business. Few countries can afford to fall behind in drone production, especially if they play a major role in technological innovation. More importantly, the civilian market for drones and their constitutive technologies is much larger than for most other dual-use goods. Nuclear reactors can be used to build bombs or to supply nuclear power (Fuhrmann 2012); missile components and surveillance satellites can be used to orchestrate remote strikes or in space programs and communications infrastructure (Mistry 2003); deadly chemicals can be weaponized or used in agriculture (Tucker 2012b). These are dual-uses, but in each case the range of military and civilian applications as well as the number of actors involved in their use is fairly small. The high cost and specialized technical knowledge required impose some limits on the number of states that could potentially develop WMDs (Tannenwald 2007, Dong-Joon and Gartzke 2007), while the narrow range of civilian applications are amenable to government monitoring and control, both at domestic and international levels (Hampson and Hart 1995; Levi and O’Hanlon 2005). Levi and O'Hanlon (2005, 2) argue that during the Cold War, arms control negotiations were facilitated by personal relationships between leaders that “could be useful for calming nerves and easing communication.” Even in an era of expanding proliferation, arms control initiatives continue to be led by high profile government administrators and work through intergovernmental organizations like the Missile Technology Control Regime, the Australia Group, and the Nuclear Suppliers Group. Drones pose special challenges for dual-use regulations because of the vast array of jobs they can perform and the much lower barriers against entering the market. West (2015) lists dozens of civilian applications for drones. Among the many tracking and inspection roles are monitoring oil and gas pipelines, utility networks, bridges, buildings, farms, and animal populations. She finds examples of them being used to fertilize crops, assist in construction projects, create maps, take photographs, and assist reporters, noting that “Some entrepreneurs have tried out drones to deliver beer at concerts and champagne to hotel balconies” (West 2015). Drones can also be purchased for personal use by hobbyists around the world, with some models available for under a hundred dollars and more advanced machines available for around a thousand dollars. The US Federal Aviation Agency (2016) predicts that the United States alone will have around 2.7 million commercial drone aircraft and 4.3 million hobbyist drone aircraft by 2020—figures that are all the more impressive considering the agency's efforts to impose limits on drone use. With such a large and diverse array of civilian uses, it is difficult to imagine building an international consensus to halt drone innovation. Even some of the most ardent activists opposing the creation of new types of lethal drones may have trouble with broad restrictions. Some of the leading figures in groups like The Campaign to Stop Killer Robots are scientists and inventors who favor nonmilitary drone use. For example, Elon Musk campaigns against lethal autonomous military drones while, at the same time, advancing complementary drone and artificial intelligence technologies via his efforts to create remote space exploration vehicles and autonomous cars (Yuhas 2016). Steve Wozniak also opposes the development of autonomous military drones, though he is positioning himself as a leader in creating autonomous cars, visual recognition devices, and civilian UAVs, which could feed into the design of military weapons platforms (Charlton 2015). Thus, the same actors that are pushing for more restraint may be obstacles against comprehensive drone controls that are broad enough to be effective against dual-use goods. The spread of civilian drones and their use by a range of violent actors demonstrates that claims about drone implementation having natural boundaries (Zenko and Kreps 2014; Gilli and Gilli 2016) are only plausible when applied to the most sophisticated military UAVs. Civilian drones are apt to be less effective in combat than those designed for military use, but their prevalence in recent and ongoing military operations provides ample evidence that drone warfare can have low entry costs and that these will continue to drop as the civilian drone market expands. The diversity of actors involved in developing drones and their heterogeneous interests raise special problems for the restriction of drones beyond those affecting other dual-use technologies and will pose major obstacles for regulatory efforts. The Cultural Dimensions of Regulation The civil-military links associated with dual-use technologies are best captured with attention to material and ideational processes. The former highlight how technologies and expertise are shared across civilian and military domains, accelerating innovation in each while also potentially drawing them closer together. The latter call attention to the cultural manifestations of “how ‘dual-use’ technologies blur the distinction between military and civilian spheres” (Kaplan, Loyer, and Daniels 2013). Culture can also be appropriated directly. For example, Benjamin reports that “those deeply involved in the military's UAV programs themselves say appealing to youth gaming culture was one of their explicit goals” when it came to attracting skilled recruits (Benjamin 2012, 86). These material and ideational senses of dual-use technologies can combine and reinforce each other, with material links promoting militarism and militarism heightening enthusiasm for powerful new weapons systems (Stahl 2009). One of the special dual-use challenges when it comes to drones is that their proliferation in nonmilitary security settings, and in civilian life more generally, could foster greater trust in drones and increase support for arming them. Existing drones depend on Global Positioning Systems (GPS), control interfaces, and telecommunications infrastructures that already show such close links between the military and civilian domains that they feature as central examples in much of the recent research on how civil-military interactions trigger cultural shifts. Kaplan et al. argue that the popularity of GPS in civilian devices contributes to the “myth of precision” in American bombing. “GPS originated as a ‘dual-use’ technology to serve both military and civilian needs. But the military needs structured the design and implementation of GPS to such a degree that everyday use brings the non-military user into militarized practices” (Kaplan et al. 2013). As evidence of how GPS alters civilian life, they trace the use of GPS to monitor children, unfaithful spouses, criminals, commercial vehicles, and goods. In each instance GPS provides the technological foundations for a mode of security that depends on assiduously tracking people's movements. The result is that “people who have no particular interest in military projects or nationalism may find themselves through their use of technology in everyday life participating in the culture of war” (Kaplan et al. 2013). Graham builds on that study, noting that “few bother to consider the ways in which military and imperial power infuse every GPS application. Every civilian use of GPS, for example, is based on US Air Force atomic clocks and geostationary satellites used to continually kill as part of the permanent, distanciated ‘war on terror’”(Graham 2012, 142). Stahl argues that the obsession with military technology is creating “techno-fetishism” that celebrates weapons advancement as an end in itself and as a symbolic mark of legitimacy (Stahl 2009). He traces various ways that civilian entertainment ranging from extreme sports to reality television helps to inculcate a “virtual citizen-soldier” attitude among civilians, which promotes militarism and the development of new technologies of war. Some of these criticisms related to the ostensible harms of civil-military technology sharing overstate the case against drones and dual-use technologies more generally. Nevertheless, they show that a complex web of material and cultural links connect the military and civilian worlds when it comes to drone operations and that these could plausibly foster greater trust in drones that would threaten support for restrictions. Commentators who advance regulatory proposals generally acknowledge the importance of promoting complementary norms and values. Boyle (2013, 28) says that “A final, and crucial, step towards mitigating the strategic consequences of drones would be to develop internationally recognized standards and norms for their use and sale.” Zenko and Kreps (2014) argue that Europe has found it much easier to control the development of UAVs because of domestic pressure. They credit opposition with discouraging Germany and other EU member states from emulating the United States’ aggressive use of UAVs. Warnings about the dangers of armed drones draw much of their strength from the unfamiliarity of these machines. Critics argue that drones could lower the threshold for initiating wars, undermine democratic accountability, and lead to more attacks on civilians (Enemark 2013, Kaag and Kreps 2014; Chamayou 2015). These claims depend heavily on speculations about possible futures in which the effects of this nascent technological revolution are felt in the most terrible ways imaginable. Uncertainty generally increases risk aversion and opposition to new technologies, while familiarity promotes trust and confidence (Nelson 2011, 131–58). Experience with self-driving cars or pizza delivery UAVs in civilian life could build confidence in drones and translate into higher estimations of the efficacy of similar machines that would be deployed in wars. Even some critics of military drones acknowledge this possibility. Zenko and Kreps (2014, 14) say that “commercial drone applications advertised by companies such as Amazon give the illusion of a technology that is ubiquitous and inevitable.” A shift in values instigated by the prevalence of drones in civilian life could undermine the consensus the opponents of drone warfare are attempting to build, and the chance of this occurring increases as drones become more common in domestic settings. We can see growing civilian comfort with the idea of expanding the use of drones from how they have been used for border security. Activists of the American Border Patrol, a civilian militia group, helped to push Customs and Border Protection towards using drones in this capacity by showing that drones could patrol the border and advocating for their use by the government (Bunn 2003). Activists in Europe have pursued a similar approach of mounting drone operations to pressure policymakers to make more extensive use of these machines. Christopher Catrambone's Migrant Offshore Aid Station (MOAS) patrols the Mediterranean searching for migrants whose boats are at risk of sinking during the difficult journey to Europe, relying on drones to provide aerial surveillance. Each migrant saved demonstrates the efficacy of using drones for humanitarian missions, and MOAS claims to have saved over fourteen thousand people as of June 2016 (Scott 2016). This kind of advocacy makes drones appear more trustworthy and demonstrates that civilian constituencies favoring greater reliance on drones already exist. Regulation and the Threat of Autonomous Drones The debate over armed drones is heavily informed by how these machines have been used by the United States and United Kingdom throughout the War on Terror, which is why commentators tend to focus on UAVs and on targeted killing at the expense of other types of drones and their use in different military operations. It remains unclear what new types of drones will be deployed in future conflicts, how they will be used, and what roles their operators will have. Disagreements are therefore guided by speculations about what the future may hold without much experience with drone warfare to draw from. The predictive task is compounded by rapid advancements and continual shifts in the direction of innovation. Because they span military and civilian applications, it is not enough for researchers to anticipate the next step in military innovation. We must also look at how the markets for drones in nonmilitary security and civilian roles are changing. Uncertainty about the future provides prudential grounds for taking a cautious approach to drones—either banning them or slowing the pace of research and development. However, the same uncertainty that helps to drive enthusiasm for regulations also points to the challenge of clearly defining their scope. Effective drone governance is apt to require prohibitions against machines that do not yet exist and with the goal of ensuring that they never come into being. The predictive character of the drone debate, especially when it comes to controlling future innovations, means that efforts to understand these machines in the present are rooted in forecasting abilities even as these forecasting abilities are hindered by myriad uncertainties and multiple different developmental trajectories across drones’ dozens of military and civilian applications. Restrictions therefore take on the future-orientation of preventing weapons technologies from being created, which is far more ambitious than the usual regulatory goal of limiting the proliferation of existing goods. Nowhere is the challenge of anticipatory regulation clearer than when it comes to semi-autonomous and autonomous drones, which are in development by several armed forces but have not been used in combat (Weinberger 2014; Baraniuk 2017). Whereas most restrictions on dual-use goods are directed at preventing the proliferation of weapons that already exist, efforts to curtail drone autonomy are aimed at preventing the creation of artificial intelligence systems with capacities that are largely speculative. Nevertheless, such regulations are attractive to many commentators because machine autonomy dramatically heightens the moral concerns associated with drones. If piloted drones create unfair risk asymmetries, lower inhibitions against using military force, and violate traditional norms of armed conflict, then semi- and fully autonomous drones exacerbate these problems by solidifying existing advantages in military technologies, making it easier to fight “riskless wars,” and reducing human control over the use of force. Autonomous drones also raise dozens of new moral and practical challenges, especially when it comes to the issue of whether these machines would be able to correctly identify targets and whether they would be susceptible to hacking or malfunctions (Sparrow 2007; Arkin 2009; Krishnan 2009; Tonkens 2013; Schulzke 2013; Johnson and Axinn 2013). Furthermore, when it comes to autonomous drones, the dual-use issues extend beyond the hardware that is typically the focus of regulatory efforts to include software that may be easier to transfer and more difficult to monitor. Opposition to autonomous drones is widespread among scientists, business leaders, academics, journalists, and activists. Icons of technological innovation, including Elon Musk, Stephen Hawking, and Steve Wozniak, issued a letter through the Future of Life Institute demanding the prohibition of autonomous weapons (Risen 2015). The Campaign to Stop Killer Robots and International Committee for Robot Arms Control host conferences and protests to spread awareness about the risks posed by the proliferation of autonomous weapons and to call for their complete prohibition. Noel Sharky (2008a, 2008b, 2010), a representative of these organizations, has issued repeated warnings about the potential dangers of an autonomous robot arms race that would lead humans to be excluded from military decision-making. Faced with this future, he contends that banning these machines is the only way to achieve some measure of security. Garcia (2014) invokes a future in which wars resemble the Terminator movies and argues that this dystopia is more realistic than it may appear to be. She says that autonomous weapons are inherently immoral and that “the United States should stand up and tell the world that people must remain in complete control when it comes to war and peace” (Garcia 2014). Although the policy recommendations remain vague, Garcia follows roughly the same path as others in suggesting that regulation is both possible and necessary. Sparrow (2016) argues that the campaign to ban autonomous weapons is akin to earlier struggles against chemical weapons and cluster bombs. He contends that a ban is essential because autonomous weapons systems are “profoundly disrespectful of the humanity of our enemy” (Sparrow 2016, 110). Wallach and Allen (2013) say that the American president should issue an executive order limiting autonomous weapon development and setting a moratorium on further steps toward their creation for at least ten years. This would then set the stage for an international effort to build a regulatory consensus. O'Connell (2013, 224) calls for a treaty to formally prohibit autonomous weapons “requiring that human beings make any decision for the near-time application of offensive lethal force to other human beings and property.” Asaro (2012) maintains that international law already provides a framework for thinking that belligerents cannot justifiably outsource the task of killing to machines and advocates their prohibition on this basis. Commentators also propose selective bans on drones that would come into effect depending on how drones are armed, where they operate, and what safety protocols they obey. Altmann (2013) would forbid those that could carry nuclear weapons or operate in space, as well as those that are too small to be visually detected. Krishnan (2009) favors a ban on where the machines can be used and how they may be armed, as well as insisting that they be fitted with kill switches. The dual-use problem with respect to autonomous systems is already evident from the rush to produce self-driving cars. Most major automotive companies are testing and selling cars that can drive themselves, and they are making major advancements in artificial intelligence systems in the process (Adams 2015). Their supporters imagine a future of dramatically reduced road traffic fatalities as human drivers are replaced by machines. Such a change would alter the experience of driving, making commutes and long road trips less monotonous and possibly even obviating onerous tasks like finding a parking place or serving as a designated driver during a night of drinking with friends. They are also amassing expertise in the design of artificial intelligence systems that could be channeled into military uses, just as other peaceful industries like nuclear power and chemical engineering may lay the foundation for weapons development programs that incorporate civilian expertise and infrastructure (Fuhrmann 2012; Tucker 2012b). Among the most interesting innovations in autonomous machines is the COTSbot, an underwater drone developed by researchers at the Queensland University of Technology (Platt 2016). It is an autonomous submarine that is meant to alleviate the problem of starfish destroying the Great Barrier Reef. This powerful new weapon in environmentalists’ arsenal would be able to kill an estimated two hundred starfish in eight hours with a pneumatic arm that injects them with bile salt. The COTSbot is not a military drone, and the bile salt injection system is clearly not the kind of weapon that could have a military application. Nevertheless, the underlying software suggests military possibilities. One of the greatest obstacles in creating lethal autonomous drones for the military is the challenge of ensuring that they select and attack the correct targets. Those who support these machines argue that they could improve compliance with the principles of discrimination and proportionality by employing strict rules of engagement and only attacking enemies whose faces are in a computer database or who are verified as hostile based on sophisticated algorithms that would be more reliable than decision procedures applied by human soldiers (Arkin 2009). Opponents of autonomous drones contend that targeting systems cannot live up to these hopes and that machines will be less accurate killers (Tonkens 2012). Whether autonomous drones would be able to select the correct targets is an empirical question that we cannot truly answer until there are machines that can actually make these decisions, but developing machines to test this proposition would immediately mark a victory for the pro-drone position. Merely testing autonomous drones to see whether they could be permissible may be a step too far based on most regulatory proposals. However, lethal civilian machines that kill nonhuman targets offer an opportunity for testing the kinds of control systems that could one day be used by armed forces. Although the weapon employed by the COTSbot may be militarily ineffective, it relies on a targeting system that closely resembles those of speculative autonomous military drones. It even includes protocols for handling ambiguous targets, as the machine can send pictures to human overseers for guidance about the right course of action when there is uncertainty. As these procedures are tested and refined they will provide invaluable insight for weapons designers interested in autonomous targeting software that could be used against humans. Like autonomous car design, this foray into lethal autonomous systems provides a forum for developing expertise in software that could be incorporated into the production of drones for military purposes. Conclusion: Reorienting Security Research The many dual-use challenges I have covered indicate that ambitious projects of completely stopping drone proliferation, imposing strong restrictions on the use of military drones, or preventing the creation of autonomous military drones are likely to fail because of the opportunities for developing drones in other domains. The dual-use character of finished drones, their constitutive parts, and their software makes it too easy to evade regulations via civilian drone research programs or to simply arm nonmilitary drones. This leaves a choice between two regulatory possibilities: broadening the proposed restrictions to cover civilian drones that have the potential to serve military ends or accepting the spread of drones across both domains and refocusing attention on how the adverse effects may be contained. The former alternative is not attractive. Expansive restrictions on innovation would seriously impede areas in which drones are making a valuable contribution (especially in humanitarian roles) and would be difficult to enforce because of the benefits of defection for states and corporations. It is also doubtful whether states would be willing to bear the economic costs of strict regulations that interfere with the civilian market. I contend that the most profitable route forward is to accept that drones are an ineliminable part of war and civilian life and to reorient research based on this starting position. This transforms the research agenda from one of formulating plans for controlling drones and their use, to one of exploring how the adverse effects of drone proliferation may be managed. Of particular importance is giving more thought to how existing institutions may limit drones’ destructiveness and slow their diffusion so that drones do not outpace the creation of new security technologies that can limit the risks of misuse. This requires devoting more attention to the mechanisms that previous work on military innovation and military balancing have uncovered as moderating influences on the disruptive impact of new technologies. Advocates of restricting drones worry that these machines could lead to a host of novel security concerns, such as lowering the threshold for initiating wars, facilitating covert operations, and making it easier for elected leaders to escape accountability for use of force decisions (Kreps and Kaag 2012; Sauer and Schörnig 2012; Enemark 2013; Kaag and Kreps 2014; Chamayou 2015). These worries are grounded in the potential for drones to create extreme asymmetries of risk that would permit states using them to circumvent the usual domestic and international barriers against launching attacks. Some critics even see asymmetry as being inherently morally objectionable (see Schulzke 2014). This suggests that the risk is not drone proliferation per se but proliferation that disrupts the military balance of power by increasing the destructiveness of attacks, advancing faster than the defensive countermeasures, or otherwise dramatically changing the security landscape. Shifting the focus of research from restricting drones to processes that could mitigate their adverse effects also provides opportunities for recognizing some advantage of the dual-use challenges I have raised. Unlike many dual-use goods that are heavily regulated, drones by themselves are unlikely to seriously threaten global security. Failing to limit the spread of WMDs, nuclear weapons, and ballistic missiles could upset the international balance of military power. If terrorist or insurgent groups were to acquire these weapons, the results could be catastrophic. The same is not true of drones. Drones are only as destructive as the weapons they are armed with, and so far, armaments have been light compared to conventional aircraft. We should be less concerned with the availability of drones and more with what weapons they are able to carry. Fortunately, regulatory regimes have had considerable success in constraining the spread of WMDs (Fuhrmann and Lupu 2016; Tucker 2012b) and ballistic missiles (Mistry 2003). Studies of drone proliferation should treat efforts to limit the spread of WMDs as one route through which drones’ adverse effects may be managed. Further investment in existing WMD controls would help to ensure that drones do not become substantially more destructive even as they become more widespread and advanced. Taking on this aspect of the dual-use problem is more likely to yield policy recommendations that can gain international ascent and build on existing successes. From a research perspective this also holds the advantage of maintaining dialogue between research on drones and work relating to other dual-use goods. Drones are also potentially much easier to defend against than WMDs and ballistic missiles. Our experience of drones has come from an era in which they were largely impervious to counterattacks, but this is unlikely to persist. Luttwak argues that the most revolutionary military devices can paradoxically be the most vulnerable to enemy countermeasures because their high profile attracts competing technologies. “The greater the success of any one technological innovation, and the more drastic the reaction involved, the more likely it is that a wide variety of scientific principles will be explored in attempting to design countermeasures. That reduces the likelihood that countermeasures will be successfully anticipated” (Luttwak 2001, 30). Rivals’ military capacities could provide a more plausible constraint than regulatory efforts, but this possibility has not been adequately explored. The drones that have been adopted by violent nonstate actors like Hezbollah and Islamic State have had some degree of success, but they have relatively little killing power and can be taken down by air defense fire, aircraft, and small arms. Even US military drones have been shot down on several occasions and are vulnerable to hacking (Woods 2014; Horowitz, Kreps, and Fuhrmann 2016). For this reason, it is essential to link future research on drones to research on the processes governing military balancing decisions. Here again the problems associated with drones may have less to do with the machines themselves than with how they fit into the dynamics of military rivalry. One benefit of studying drones as dual-use technologies is that it also paves the way for recognizing that countermeasures may span military and civilian roles. Law enforcement agencies in the United States and United Kingdom in particular have been working on a range of antidrone devices, such as radio interference, special munitions, nets, and attack birds (Telegraph 2017). Even private entities fearing espionage, such as film studios and technology firms, are investing in these devices (Greenberg 2017). Just like drones themselves, defensive technologies may be imported into civilian and military roles to prevent drones from becoming serious security risks. With this in mind, future research should not only look at drones within the dual-use framework but also the potential countermeasures. From this perspective, dual-use transforms from a security problem into a framework for establishing civil-military cooperation to protect against drones. Another important area of research is into the incentives that lead violent actors to adopt drones, either by developing their own variants or adapting them from civilian or nonmilitary security roles. Proponents of controlling drone proliferation raise reasonable concerns about the future, but studies of violent organizations’ internal dynamics indicate that these fears may overstate violent actors’ goals. Dolnik (2007) argues that terrorist organizations tend to be fairly conservative when it comes to innovation and that their weapons and tactics advance incrementally. Even dynamic organizations like al Qaeda face internal resistance to change and take an incremental or evolutionary approach to adopting new technologies and practices (Jackson and Loidolt 2013). Given their limited resources, it is understandable that terrorists and insurgents may not wish to invest too heavily in untested strategies. It is imperative for researchers to look into what specific motives might be at work when it comes to leading these groups to use drones and what barriers they face. State military forces likewise tend to be conservative when it comes to embracing new technologies (Rosen 1991; Posen 1986). The US Air Force was slow to adopt drones, only arming the Predator after six years of unarmed service, and ground-based drones have yet to achieve any meaningful role in combat, despite the US military having these machines in its arsenal for more than a decade (Woods 2015). Grissom (2006, 119) finds that “the four contemporary schools of military innovation, and virtually every major study on the subject, argue that military organizations are intrinsically inflexible, prone to stagnation, and fearful of change.” Research on drone proliferation would be well-served by taking work on military innovation into account when forecasting. Institutional conservatism is apt to be an important aspect of moderating the reliance on drones, especially if regulators can understand what incentives are at work and appeal to them. Regulating drone use may not be necessary if we can understand how existing incentives toward conservatism are already encouraging moderation in drone research and implementation. Moreover, these institutional barriers are apt to persist even as drones become more common in civilian life. The course correction in research that I propose should go beyond work specifically dealing with drones to affect research on dual-use goods in general, civil-military relations, and military innovation. The study of dual-use goods should make drones and their associated technologies a central concern. For one thing, drones present unique challenges that I have identified throughout the essay, and these must form part of the overall framework for understanding what dual-use goods are and how they are used. This area of research can offer some of the best strategies for managing drone diffusion and encouraging the spread of defensive technologies when strict regulations are impossible. It is especially important for future research to explore the extent to which drones may fit within existing regulatory regimes and how controls on WMDs can be framed to help ensure that drones are not used as delivery vehicles. The dual-use problem involving drones will affect civil-military relations by undermining the distinction between these two domains. With hardware, software, and operational skills crossing between civilian and military spheres, it will be increasingly difficult to locate drone use in one domain or the other. It will be up to research on civil-military relations to map out these interactions and to measure the effects of drone proliferation alongside other factors blurring the boundaries between these domains, such as the spread of asymmetric violence. Perhaps the most valuable contribution would come from tracking the spread of civilian drones into militarized roles, so we can understand the scale of the risk civilian drones present. Given the risks associated with civil-military technology sharing, work on military innovation needs to devote far more attention to nonmilitary sources of technological advancement. The dominant frameworks for studying military innovation focus primarily, or even entirely, on causal mechanisms that operate within armed forces (see Grissom 2006). The prospect of heavy influence coming from the civilian market means that new directions of diffusion need to be mapped out. This is especially true when it comes to studying innovation by violent nonstate actors and weak states, as they are apt to be more heavily engaged with the civilian market than state military forces that have major internal research arms. As these examples indicate, the dual-use problem associated with drones has sweeping implications that affect many different areas of security research; it is not merely a barrier against drone regulation. With drone technology and proliferation moving at a rapid pace, it is imperative for scholars across specializations to engage in a collective effort to understand drones’ impact and to explore the strategies that may help to moderate the adverse costs of their proliferation. References Adams T. 2015 . “ Self-Driving Cars: From 2020 You Will Become a Permanent Backseat Driver .” Guardian , September 13, 2015. Accessed January 10, 2017. https://www.theguardian.com/technology/2015/sep/13/self-driving-cars-bmw-google-2020-driving . WorldCat Altman J. 2013 . “ Arms Control for Armed Uninhabited Vehicles: An Ethical Issue .” Ethics and Information Technology 15 (2) : 137 – 52 . Google Scholar Crossref Search ADS WorldCat Arkin R. C. 2009 . Governing Lethal Behavior in Autonomous Robots . Boca Raton, FL : Taylor & Francis Group . Google Preview WorldCat COPAC Asaro P. 2012 . “ On Banning Autonomous Weapon Systems: Human Rights, Automation, and the Dehumanization of Lethal Decisionmaking .” International Review of the Red Cross 94 ( 886 ): 687 – 709 . Google Scholar Crossref Search ADS WorldCat Baraniuk C. 2017 . “US Military Tests Swarm of Mini-Drones Launched from Jets.” BBC News , January 10, 2017. http://www.bbc.co.uk/news/technology-38569027 . BBC . 2007 . “Pilotless Police Drone Takes Off.” BBC News 21 May, 2007. Accessed January 10, 2017. http://news.bbc.co.uk/1/hi/england/merseyside/6676809.stm .. Benjamin M. 2012 . Drone Warfare: Killing by Remote Control . New York : OR Books . Google Preview WorldCat COPAC Bertsch G. K. , Grillot S. R. , ed. 1998 . Arms on the Market: Reducing the Risk of Proliferation in the Former Soviet Union . New York : Routledge . Google Preview WorldCat COPAC Bodeti A. 2016 . “ Iraqi Militias Are Using Consumer Drones to Fight Isis .” Wired , June 28, 2016. Accessed January 10, 2017. http://www.wired.co.uk/article/iraq-isis-war-consumer-drones . WorldCat Bolton D. 2016 . “ Domino's Set to Introduce Pizza Delivery Robots in New Zealand .” Independent , March 18, 2016. Accessed January 10, 2017. http://www.independent.co.uk/life-style/gadgets-and-tech/news/dominos-pizza-delivery-robots-new-zealand-a6939536.html . WorldCat Boyle M. J. 2013 . “ The Costs and Consequences of Drone Warfare .” International Affairs 89 ( 1 ): 1 – 29 . Google Scholar Crossref Search ADS WorldCat Buchanan A. , Keohane R. O. . 2015 . “ Toward a Drone Accountability Regime .” Ethics & International Affairs 29 ( 1 ). WorldCat Bunn A. 2003 . “ Homegrown Homeland Defense .” New York Times , June 1, 2003. Accessed January 10, 2017. http://www.nytimes.com/2003/06/01/magazine/01ENCOUNTER.html?pagewanted=all . WorldCat Busch N. E. , Joyner D. H. , ed. 2009 . Combating Weapons of Mass Destruction . Athens : University of Georgia Press . Google Preview WorldCat COPAC Byman D. 2013 . “ Why Drones Work .” Foreign Affairs July/August, 2013. Accessed January 10, 2017. https://www.foreignaffairs.com/articles/somalia/2013-06-11/why-drones-work . WorldCat Chamayou G. 2015 . A Theory of the Drone . Translated by Lloyd Janet . New York : The New Press . Google Preview WorldCat COPAC Charlton A. 2015 . “ Apple Will Have Share of Autonomous Car Market Says Co-Founder Steve Wozniak .” International Business Times June 26 http://www.ibtimes.co.uk/apple-will-have-share-autonomous-car-market-says-co-founder-steve-wozniak-1508067 . WorldCat Clutterbuck R. 1994 . Terrorism in an Unstable World . New York : Routledge . Google Preview WorldCat COPAC Corera G. 2006 . Shopping for Bombs: Nuclear Proliferation, Global Insecurity, and the Rise and Fall of the A.Q. Khan Network . New York : Hurst . Google Preview WorldCat COPAC Csernatoni R. 2016 . “ High-Tech Fortress Europe: FRONTEX and the Dronization of Border Management .” European Public Affairs January 27, 2016. Accessed January 10, 2017. http://www.europeanpublicaffairs.eu/high-tech-fortress-europe-frontex-and-the-dronization-of-border-management/ . WorldCat Della Cava Marco . 2015 . “ Police Taser Drones Authorized in N.D .” USA Today 29 August http://www.usatoday.com/story/tech/2015/08/28/police-taser-drones-authorized–north-dakota/71319668/ . WorldCat Dolnik A. 2007 . Understanding Terrorist Innovation: Technology, Tactics and Global Trends . New York : Routledge . Google Preview WorldCat COPAC Dong-Joon J. , Gartzke E. . 2007 . “ Determinants of Nuclear Weapons Proliferation .” Journal of Conflict Resolution 51 ( 1 ): 167 – 94 . Google Scholar Crossref Search ADS WorldCat Dunn D. H. 2013 . “ Drones: Disembodied Aerial Warfare and the Unarticulated Threat .” International Affairs 89 ( 5 ): 1237 – 46 . Google Scholar Crossref Search ADS WorldCat Edwards H. S. 2016 . “ When Can Police Use a ‘Bomb Robot’ to Kill a Suspect? ” Time , July 8, 2016. Accessed January 10, 2017. http://time.com/4398196/dallas-shooting-bomb-robot/ . WorldCat Enemark C. 2011 . “ Drones over Pakistan: Secrecy, Ethics, and Counterinsurgency .” Asian Security 7 ( 3 ): 218 – 37 . Google Scholar Crossref Search ADS WorldCat Enemark C. 2013 . Armed Drones and the Ethics of War: Military Virtue in a Post-heroic Age . New York : Routledge . Google Preview WorldCat COPAC Farmer B. 2015 . “ Royal Navy to Send Drones to the Mediterranean to Save Migrants .” Telegraph , 27 April, 2015. Accessed January 10, 2017. http://www.telegraph.co.uk/news/worldnews/africaandindianocean/libya/11564322/Royal-Navy-to-send-drones-to-the-Mediterranean-to-save-migrants.html . WorldCat Federal Aviation Agency . 2016 . FAA Aerospace Forecast, Fiscal Years 2016–2036 . Accessed January 10, 2017. https://www.faa.gov/data_research/aviation/aerospace_forecasts/media/FY2016-36_FAA_Aerospace_Forecast.pdf . WorldCat COPAC Fuhrmann M. 2008 . “ Exporting Mass Destruction? The Determinants of Dual-Use Trade .” Journal of Peace Research 45 ( 5 ): 633 – 52 . Google Scholar Crossref Search ADS WorldCat Fuhrmann M. 2009 . “ Taking a Walk on the Supply Side: The Determinants of Civilian Nuclear Cooperation .” Journal of Conflict Resolution 53 ( 2 ): 181 – 208 . Google Scholar Crossref Search ADS WorldCat Fuhrmann M. 2012 . Atomic Assistance: How “Atoms for Peace” Programs Cause Nuclear Insecurity . Ithaca, NY : Cornell University Press . Google Preview WorldCat COPAC Fuhrmann M. , Lupu Y. . 2016 . “ Do Arms Control Treaties Work? Assessing the Effectiveness of the Nuclear Nonproliferation Treaty .” International Studies Quarterly 60 ( 3 ): 530 – 39 . Google Scholar Crossref Search ADS WorldCat Garcia D. 2014 . “ The Case against Killer Robots: Why the United States Should Ban Them .” Foreign Affairs , May 10, 2014. Accessed January 10, 2017. https://www.foreignaffairs.com/articles/united-states/2014-05-10/case-against-killer-robots . WorldCat Gilli A. , Gilli M. . 2016 . “ The Diffusion of Drone Warfare? Industrial, Organizational, and Infrastructural Constraints .” Security Studies 25 ( 1 ): 50 – 84 . Google Scholar Crossref Search ADS WorldCat Graham S. 2012 . “ When Life Itself Is War: On the Urbanization of Military and Security Doctrine .” International Journal of Urban and Regional Research 36 ( 1 ): 136 – 55 . Google Scholar Crossref Search ADS WorldCat Greenberg A. 2017 . “ Watch a Test of Anti-drone Weapons .” Wired , July 26, 2017. Accessed January 10, 2017. https://www.wired.com/story/watch-anti-drone-weapons-test/ . WorldCat Grissom A. 2006 . “ The Future of Military Innovation Studies .” The Journal of Strategic Studies 29 ( 5 ): 905 – 34 . Google Scholar Crossref Search ADS WorldCat Hambling D. 2016 . “How Islamic State is Using Consumer Drones.” BBC News , December 9, 2016. Accessed January 10, 2017. http://www.bbc.com/future/story/20161208-how-is-is-using-consumer-drones . Hampson F. O. , Hart M. . 1995 . Multilateral Negotiations: Lessons from Arms Control, Trade, and the Environment . Baltimore, MD : John Hopkins University Press . Google Preview WorldCat COPAC Hoenig M. 2014 . “ Hezbollah and the Use of Drones as a Weapon of Terrorism .” Public Interest Report 67 ( 2 ). WorldCat Horowitz M. C. , Kreps S. E. , Fuhrmann M. . 2016 . “ Separating Fact from Fiction in the Debate over Drone Proliferation .” International Security 41 ( 2 ): 7 – 42 . Google Scholar Crossref Search ADS WorldCat Jackson B. A. , Loidolt B. . 2013 . “ Considering al-Qa'ida's Innovation Doctrine: From Strategic Texts to ‘Innovation in Practice.’ ” Terrorism and Political Violence 25 ( 25 ): 284 – 310 . Google Scholar Crossref Search ADS WorldCat Johnson Aaron , Axinn Sidney . 2013 . “ The Morality of Autonomous Robots .” Journal of Military Ethics 12 ( 2 ): 129 – 41 . Google Scholar Crossref Search ADS WorldCat Joscelyn T. 2016 . “ Jund Al Aqsa Uses Drone to Drop Small Bomb on Syrian Regime Forces .” Threat Matrix , September 2, 2016. Accessed January 10, 2017. http://www.longwarjournal.org/archives/2016/09/jund-al-aqsa-uses-drone-to-drop-small-bomb-on-syrian-regime-forces.php . WorldCat Kaag J. , Kreps S. . 2014 . Drone Warfare . Malden, MA : Polity Press . Google Preview WorldCat COPAC Kaplan C. , Loyer E. , Daniels E. C. . 2013 . “ Precision Targets: GPS and the Militarization of Everyday Life .” Canadian Journal of Communication 38 ( 3 ). Accessed January 10, 2017. http://www.cjc-online.ca/index.php/journal/article/view/2655/2721 . WorldCat Kilcullen D. , Exum A. M. . 2009 . “ Death from Above, Outrage Down Below .” New York Times , May 16, 2009. Accessed January 10, 2017. http://www.nytimes.com/2009/05/17/opinion/17exum.html?pagewanted=all&_r=0 . WorldCat Koplow D. A. 2006 . Non-lethal Weapons: The Law and Policy of Revolutionary Technologies for the Military and Law Enforcement . Cambridge : Cambridge University Press . Google Preview WorldCat COPAC Koplow D. A. 2010 . Death by Moderation: The U.S. Military's Quest for Useable Weapons . New York : Cambridge University Press . Google Preview WorldCat COPAC Kreps S. , Kaag J. . 2012 . “ The Use of Unmanned Aerial Vehicles in Asymmetric Conflict: Legal and Moral Implications .” Polity 44 ( 2 ): 260 – 85 . Google Scholar Crossref Search ADS WorldCat Krishnan A. 2009 . Killer Robots: Legality and Ethicality of Autonomous Weapons . Burlington, VT : Ashgate . Google Preview WorldCat COPAC Levi M. A. , O'Hanlon M. E. . 2005 . The Future of Arms Control . Washington, DC : The Brookings Institution . Google Preview WorldCat COPAC Lin J , Singer P. W. . 2016 . “ China Debuts the AnBot, the Police Robot .” Popular Science April 27, 2016. http://www.popsci.com/china-debuts-anbot-police-robot . WorldCat Luttwak E. N. 2001 . Strategy: The Logic of War and Peace . Cambridge, MA : Belknap Press . Google Preview WorldCat COPAC Meier P. 2015 . Digital Humanitarians: How Big Data Is Changing the Face of Humanitarian Response . New York : CRC Press . Google Preview WorldCat COPAC Mistry D. 2003 . Containing Missile Proliferation: Strategic Technology, Security Regimes, and International Cooperation in Arms Control . Seattle : University of Washington Press . Google Preview WorldCat COPAC Nelson L. S. 2011 . America Identified: Biometric Technology and Society . Cambridge, MA : MIT Press . Google Preview WorldCat COPAC Nixon R. 2016 . “ Drones, So Useful in War, May Be Too Costly for Border Duty .” New York Times , November 2, 2016. Accessed January 10, 2017. https://www.nytimes.com/2016/11/03/us/drones-canadian-border.html . WorldCat O'Connell M. E. 2013 . “ Banning Autonomous Killing .” In The American Way of Bombing: How Legal and Ethical Norms Change , edited by Evangelista Matthew , Shue Henry . Ithaca, NY : Cornell University Press . Google Preview WorldCat COPAC Platt J. R. 2016 . “ A Starfish-Killing, Artificially Intelligent Robot Is Set to Patrol the Great Barrier Reef .” Scientific American , January 1, 2016. Accessed January 10, 2017. http://www.scientificamerican.com/article/a-starfish-killing-artificially-intelligent-robot-is-set-to-patrol-the-great-barrier-reef/ . WorldCat Posen B. R. 1986 . The Sources of Military Doctrine . Ithaca, NY : Cornell University Press . Google Preview WorldCat COPAC Risen T. 2015 . “ Musk, Hawking Call to Ban Robots That Could Kill You All By Themselves .” US News & World Report , July 27, 2015. Accessed January 10, 2017. http://www.usnews.com/news/articles/2015/07/27/elon-musk-stephen-hawking-fight-to-ban-autonomous-weapons . WorldCat Rosen S. P. 1991 . Winning the Next War: Innovation and the Modern Military . Ithaca, NY : Cornell University Press . Google Preview WorldCat COPAC Sauer F. , Schörnig N. 2012 . “ Killer Drones: The ‘Silver Bullet’ of Democratic Warfare? ” Security Dialogue 43 ( 4 ): 363 – 80 . Google Scholar Crossref Search ADS WorldCat Schulzke Marcus . 2011 . “ Robots as Weapons in Just Wars .” Philosophy and Technology 24 ( 3 ): 293 – 306 . Google Scholar Crossref Search ADS WorldCat Schulzke M. 2013 . “ Autonomous Weapons and Distributed Responsibility .” Philosophy and Technology 26 ( 2 ): 203 – 19 . Google Scholar Crossref Search ADS WorldCat Schulzke M. 2014 . “ The Morality of Remote Warfare: Against the Asymmetry Objection to Remote Weaponry .” Political Studies 64 ( 1 ): 90 – 105 . Google Scholar Crossref Search ADS WorldCat Schulzke M. 2017 . The Morality of Drone Warfare and the Politics of Regulation . London : Palgrave . Google Preview WorldCat COPAC Scott K. 2016 . “ Thousands of Migrants Are Dying at Sea. This Charity is Trying to Save Them .” Wired , June 13, 2016. Accessed January 10, 2017. http://www.wired.co.uk/article/app-save-refugees-moas . WorldCat Shane S. 2015 . Objective Troy: A Terrorist, a President, and the Rise of the Drone . New York : Tim Duggan Books . Google Preview WorldCat COPAC Sharkey N. 2008a . “ Cassandra or False Prophet of Doom: AI Robots and War .” Intelligent Systems 23 ( 4 ): 14 – 17 . Google Scholar Crossref Search ADS WorldCat Sharkey N. 2008b . “ The Ethical Frontiers of Robotics .” Computer Science 322 ( 5909 ): 1800 – 1 . WorldCat Sharkey N. 2010 . “ Saying ‘No!’ to Lethal Autonomous Targeting .” Journal of Military Ethics 9 ( 4 ): 369 – 83 . Google Scholar Crossref Search ADS WorldCat Singer P. W. 2009 . Wired for War: The Robotics Revolution and Conflict in the 21st Century . New York : Penguin Press . Google Preview WorldCat COPAC Singer P. W. 2015 . “ The Future of War Will Be Robotic .” CNN , February 23, 2015. Accessed January 10, 2017. http://edition.cnn.com/2015/02/23/opinion/singer-future-of-war-robotic/ . WorldCat Sparrow R. 2007 . “ Killer Robots .” Journal of Applied Philosophy 24 ( 1 ): 62 – 77 . Google Scholar Crossref Search ADS WorldCat Sparrow R. 2016 . “ Robots and Respect: Assessing the Case Against Autonomous Weapon Systems .” Ethics & International Affairs 30 ( 1 ): 93 – 116 . Google Scholar Crossref Search ADS WorldCat Stahl R. 2009 . Militainment, Inc.: War, Media, and Popular Culture . New York : Routledge . Google Preview WorldCat COPAC Strawser B. J . 2010 . “ Moral Predators: The Duty to Employ Uninhabited Aerial Vehicles .” Journal of Military Ethics 9 ( 4 ): 342 – 68 . Google Scholar Crossref Search ADS WorldCat Stinnett D. M. , Early B. R. , Horne C. , Karreth J. . 2011 . “ Complying by Denying: Explaining Why States Develop Nonproliferation Export Controls .” International Studies Perspectives 12 ( 3 ): 308 – 26 . WorldCat Tannenwald N. 2007 . The Nuclear Taboo: The United States and the Non-use of Nuclear Weapons Since 1945 . Cambridge : Cambridge University Press . Google Preview WorldCat COPAC Telegraph . 2017 . “ British Prison is First to Use ‘Disruptor’ to Create Drone-Proof ‘Shield’ Around Jail .” Telegraph , May 16, 2017. Accessed January 10, 2017. http://www.telegraph.co.uk/news/2017/05/16/british-prison-first-use-disruptor-create-drone-proof-shield/ . WorldCat Tonkens R. 2012 . “ The Case Against Robotic Warfare: A Response to Arkin .” Journal of Military Ethics 11 ( 2 ): 149 – 68 . Google Scholar Crossref Search ADS WorldCat Tonkens R. 2013 . “ Should Autonomous Robots Be Pacifists .” Ethics and Information Technology 15 : 109 – 23 . Google Scholar Crossref Search ADS WorldCat Tucker J. B. 2012a . “ Review of the Literature on Dual Use .” In Innovation, Dual Use, and Security: Managing the Risks of Emerging Biological and Chemical Technologies , edited by Tucker Jonathan B. , 19 – 44 . Cambridge, MA : MIT Press . Google Preview WorldCat COPAC Tucker J. B. , ed. 2012b . Innovation, Dual Use, and Security: Managing the Risks of Emerging Biological and Chemical Technologies . Cambridge, MA : MIT Press . Google Preview WorldCat COPAC Vanian J. 2016 . “ Amazon's Drone Testing Takes Flight in Yet Another Country .” Fortune , February 1, 2016. Accessed January 10, 2017. http://fortune.com/2016/02/01/amazon-testing-drones-netherlands/ . WorldCat Waddell K. 2015 . “ Few Privacy Limitations Exist on How Police Use Drones .” Atlantic , February 5, 2015. Accessed January 10, 2017. http://www.theatlantic.com/politics/archive/2015/02/few-privacy-limitations-exist-on-how-police-use-drones/458583/ . WorldCat Wallach W , Allen C. . 2013 . “ Framing Robot Arms Control .” Ethics and Information Technology 15 ( 2 ): 125 – 35 . Google Scholar Crossref Search ADS WorldCat Weinberger S. 2014 . “ X-47B Stealth Drone Targets New Frontiers .” BBC News , November 18, 2014. Accessed January 10, 2017. http://www.bbc.com/future/story/20121218-stealth-drone-targets-life-at-sea . WorldCat West G. 2015 . “ Drone On: The Sky's the Limit—If the FAA Will Get Out of the Way .” Foreign Affairs , May/June, 2015. Accessed January 10, 2017. https://www.foreignaffairs.com/articles/2015-05-01/drone . WorldCat Whittle R. 2014 . Predator: The Secret Origins of the Drone Revolution . New York : Henry Holt . Google Preview WorldCat COPAC Wired . 2016 . “ Surrey Now Has the UK's ‘Largest’ Police Drone Project .” Wired April 12, 2016. Accessed January 10, 2017. http://www.wired.co.uk/news/archive/2016-04/12/surrey-police-uk-largest-drone-trial . WorldCat Woods C. 2014 . Sudden Justice: The True Costs of America's Secret Drone War . London : C Hurst & Company . Google Preview WorldCat COPAC Woods C. 2015 . “ The Story of America's First Drone Strike .” Atlantic May 30, 2015. Accessed January 10, 2017. http://www.theatlantic.com/international/archive/2015/05/america-first-drone-strike-afghanistan/394463/ . WorldCat Yuhas A. 2016 . “ SpaceX Rocket Successfully Lands on Ocean Drone Platform for First Time .” Guardian , April 8, 2016. Accessed January 10, 2017. https://www.theguardian.com/science/2016/apr/08/spacex-rocket-test-elon-musk-international-space-station . WorldCat Zenko M. , Kreps S. . 2014 . Limiting Armed Drone Proliferation . New York : Council of Foreign Relations . Google Preview WorldCat COPAC © The Author(s) (2018). Published by Oxford University Press on behalf of the International Studies Association. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Drone Proliferation and the Challenge of Regulating Dual-Use Technologies
JF - International Studies Review
DO - 10.1093/isr/viy047
DA - 2019-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/drone-proliferation-and-the-challenge-of-regulating-dual-use-zRCcxAV2ZR
SP - 497
VL - 21
IS - 3
DP - DeepDyve
ER -