on march 11, 2011, a massive underwater earthquake struck approximately 43 miles from the northeastern coast of Honshu, the largest island in Japan. Registering 9.0 on the Richter scale, this earthquake was the fourth largest recorded since 1900.1 It generated a ferocious tsunami with waves as high as 127 feet2 that rapidly devastated the eastern coast of Japan, killing 15,890 people and leaving another 2,590 missing and presumed dead. While it caused over $220 billion in damage in Japan alone, its ripple effects were felt thousands of miles away in Indonesia, Papua New Guinea, Hawaii, and even California. What made this event particularly devastating was the Fukushima nuclear disaster that followed it. The Fukushima Daiichi nuclear power plant—a six-reactor nuclear plant first commissioned in 1971—lay directly in the path of the tsunami. Although the reactors at Fukushima shut down as soon as the earthquake struck, the emergency generators cooling the reactors were badly damaged. In the days that followed, three nuclear reactors melted down and radioactive material began to leak into the surrounding area. Approximately 300,000 people were rapidly evacuated from the disaster zone. Whole towns and villages were left as ghostly reminders of the life that existed there before the meltdown. The surrounding region was found to be so radioactive that the Japanese government declared it an exclusion zone, effectively rendering it off limits for decades.
Following the disaster, the Japanese government came under harsh scrutiny for inattentive oversight and regulation and poor communication during the crisis. An independent inquiry later revealed that Japanese officials urged calm while secretly fearing a “demonic chain reaction” through which the Fukushima radiation would affect other nearby nuclear reactors, leading to cascading crises and the eventual abandonment of Tokyo.3 Facing protests and calls to forever end Japan’s reliance on nuclear power, the Japanese government had temporarily shut down all fifty of its nuclear reactors for maintenance and safety inspections by 2012. It also committed billions to decommissioning the reactors and to the decades-long environmental clean-up. Yet it would not bow to the protests and end Japan’s reliance on nuclear power. In 2014, a new government, dominated by the Liberal Democratic Party (LDP) and led by Prime Minister Shinzo Abe, made it official policy to embrace nuclear power as part of its balanced energy plan. By 2015, reactors were coming back online despite a vocal protest movement that no longer trusted the Japanese government to conduct proper and safe oversight of these facilities.
Among these critics was Yasuo Yamamoto, a forty-year-old unemployed man who lived in the town of Obama in Fukui prefecture in western Japan. On April 9, 2015, Yamamoto flew a small DJI Phantom 2 drone, one of the most popular commercial models available, onto the rooftop of Prime Minister Abe’s office. The drone was painted black and marked with a small red symbol for radioactive material. It carried a camera and a small plastic bottle containing radioactive sand from the Fukushima region. The amount of radioactive material, cesium, was ultimately found to be so small that it was unlikely to affect humans or the environment.4 But this event marked the first time that anyone had gotten radioactive material that close to a world leader. It was also not quickly discovered: the drone sat on the rooftop of the prime minister’s office for two weeks until it was discovered by an employee giving tours to visitors.5 If the drone had carried more dangerous radioactive materials, the threat to lives of Prime Minister Abe and others in the area could have been much worse.
On April 25, Yamamoto turned himself into the police and admitted that the drone incursion was designed to protest the government’s decision to restart Japan’s nuclear reactors. The point was clear: if the Japanese government was going to restart nuclear power plants in his prefecture—the location of nearly one quarter of all Japan’s reactors—then Yamamoto was going to bring those risks home to those making the decision.6 After his arrest, a series of blog posts by Yamamoto were discovered in which he showed off the next drone he planned to launch and remarked that “in order to prevent the restarting of nuclear reactors, one cannot rule out terrorism.”7
Although Yamamoto’s protest did not cause any direct harm to Prime Minister Abe or anyone else, it provided an alarming illustration of how terrorists might use drones for assassinations and even terrorist attacks. The Japanese government reacted swiftly by banning drones over high-profile government sites and nuclear reactors, and later it enacted a wider ban of drone flights over most of Tokyo. It also began researching a number of different technologies—including radar and sonar detection, as well as police drones deploying nets to capture suspicious drones—to protect critical sites in Japan.8 But none of these measures could wholly eliminate the risk of a drone attack by a disgruntled person or group. Yamamoto’s “drone protest” showed that a small commercial drone packed with explosives or radioactive material could access well-defended sites, and even possibly spread radioactive material, without notice or clear attribution.
The potentially grave risks associated with terrorist drones were highlighted by President Barack Obama at the Nuclear Security Summit in Washington, DC, in April 2016. At a meeting of fifty heads of state, President Obama sketched a scenario where radioactive material was purchased on the so-called dark web—the uncharted part of the internet, where illicit material is often sold and exchanged with little scrutiny—and attached to drones to launch attacks in Western cities.9 His point was that if governments could not control drone diffusion, they should at least try to control access to radiological material. Although President Obama’s “dirty drone” scenario was hypothetical, British Prime Minister David Cameron warned that the threat of a group like the Islamic State deploying dirty bomb drones was “only too real.” British intelligence officials also revealed that the Islamic State had been seeking low-level crop duster drones which would be ideal for spreading radioactive material.10Given that radiological material exists in governments and research labs in 130 countries, and that drones are spreading like wildfire across the world, some officials worry that it is only a matter of time before a drone-based terrorist attack, possibly involving nuclear material, becomes a reality in a Western city. In September 2017, FBI director Christopher Wray testified that drone terrorism is “coming here, imminently.”11
While some of these dire predictions may be overblown, it is clear that the drone age is one in which non-state actors—such as terrorist groups, but also rebel groups fighting for different reasons—will embrace this technology for their own ends. To some extent, this was inevitable. Given how easy it is to acquire and fly a drone, and the difficulty in enacting surveillance over all potential targets of attack, it would be impossible to stop non-state actors from turning drones into weapons. There is also no doubt that they would be interested: many non-state actors, especially terrorists, are naturally risk-taking and would have no compunction about using drones to advance their cause. But the ultimate effect of the ability of non-state actors to access drone technology remains unclear in two respects. First, will drones expand the goals of these organizations? Will terrorists or rebel armies choose to do more or do things differently as a result of having access to drones? Second, will the introduction of drones make a difference? The wars between governments and non-state actors—for example, the United States versus the Islamic State—are typically marked by stark asymmetries in power, with the governments having a decisive advantage in most respects. While it is possible that drones will only be an occasional threat, even a nuisance, which does not materially change the outcome of the battle between them, it is also possible that drone technology may enable non-state actors to begin to level the playing field against their more powerful opponents in new and surprising ways.
To answer these questions, it is first important to understand what non-state actors like terrorist organizations will be able to do with drones. The most widely known military drones—for example, the Reaper and the Global Hawk—are far beyond the reach of even powerful terrorist groups like Hezbollah or the Islamic State due to their high cost and restrictions on their sale.12 But with hundreds of smaller drones available on the commercial market, there is no shortage of drones that could be used for terrorist attacks. Given this fact, governments around the world are beginning to prepare for these attacks. The US government has identified risks associated with both commercial and hobbyist drones being converted into vehicles for attack.13 In 2013, the National Counterterrorism Center convened a sixty-five-member working group tasked with analyzing the threat and preparing a response.14 In 2015, the Department of Homeland Security (DHS) issued a worldwide alert that commercially available drones might be used in attacks on crucial infrastructure, such as airports.15 Similar concerns have also been expressed in Britain, which has established a cross-government working group to address this risk.16 Some have described the current state of play as a “new global arms race” between terrorists seeking to turn drones into weapons and the governments that are struggling to develop effective countermeasures.17
One of the most common fears is that terrorist organizations will pack explosives onto a drone and drop the bombs on a populated area, causing mass casualties. A number of different types of explosives—mortars, hand grenades, and other IEDs, for example—could be dropped from modestly sized drones and injure people or damage buildings. Although it would pose a technical challenge, more than one explosive device could also be attached to a drone for maximum damage.18 Multiple drones could also be used at the same time to scout targets, drop multiple payloads, or even confuse defenders. The degree of the damage caused by such an attack is likely to vary based on the payload and type of explosive employed. A small hobbyist drone could drop a small payload—estimated to be 5–10 kg of TNT, roughly the equivalent of a pipe bomb or suicide vest—to devastating effect if the target was hit precisely.19
In an alternative scenario, terrorists turn the drones themselves into bombs. An explosive-laden drone could also be flown directly into a crowd, causing casualties and generating mass panic, or into a fixed target at a symbolic or high-profile location. There is already some evidence that these attacks have been planned. In 2010, Rezwan Ferdaus was arrested in Massachusetts for planning to slam remote-controlled planes, each containing 5 pounds of plastic explosives, into the Pentagon and US Capitol building.20 In 2014, a Moroccan national, El Mehdi Semlali Fathi, was arrested for plans to pack explosives onto radio-controlled airplanes and fly them into a school and a federal building.21 Most of these attempts were amateurish, but there is a real danger that the use of drones on the battlefield in Iraq and Syria by the Islamic State and others will produce expertise and technical adaptations needed to make such attacks feasible elsewhere. Such attacks would not need to have pinpoint accuracy to be effective; the degree of the panic they would cause would be enough to render the attack a success.
For terrorists, there are also clear benefits to deploying a large number of explosive-laden drones at the same time. The basic principles of deploying drones in coordinated arrays has been illustrated by the drone light shows at major sporting events and Disneyland. But the tactical benefits of such an approach are also compelling: multiple drones moving in tandem could overwhelm defenses, confuse defenders, and ensure that a target is struck. Similarly, swarms of drones—that is, coordinated drones which can move semi-autonomously and adaptively to hit a particular target—would be ideal for terrorist use. For example, terrorists could launch a swarm of small drones, packed with explosives, into a major sporting event like the Super Bowl.22 A swarm of drones would represent the equivalent of “multiple suicide bombers launched at a single target at the same time.”23 Swarms are particularly hard to neutralize as a threat with conventional weapons like firearms given their adaptability and persistence once one has been knocked down.24 Swarms are out of the reach for terrorist groups without a deep bench of technical expertise at present, but as the technology to create swarms becomes more common in the commercial sphere the risk of swarming drone attacks will naturally increase.
Another risk is of drone assassinations. In 2013, retired Admiral Dennis Blair, who previously served as Director of National Intelligence for President Obama, warned that explosive-laden drones could be used to kill high-level political and military figures.25 Al Qaeda has long had an interest in drone assassinations, having plotted to kill President George W. Bush with a remote-controlled airplane at the G-8 summit in Genoa, Italy, in 2001.26 The fears that such an attack might be possible were heightened in September 2013, when a protester managed to fly a small quadcopter within a few meters of Chancellor Angela Merkel and other German officials at a public event.27 Other events have demonstrated the vulnerability of world leaders to such attacks. In 2018, Saudi Arabian security forces shot down a small drone approaching a royal palace in Riyadh.28 While King Salman was not present at the time, the shooting and subsequent confusion raised the specter of a coup attempt.29 While neither of these attempts ultimately came close to harming a world leader, they showed that a drone assassination attempt was in fact possible.
Nothing illustrated this fact to the wider public more than the attempted assassination of Venezuelan president Nicholas Maduro with a drone on August 4, 2018. On a Saturday afternoon at 5:30 p.m., Maduro was giving a speech in Caracas to celebrate the eighty-first anniversary of the formation of the National Guard. This was a reasonably well-protected speech: the embattled leader, blamed for the country’s economic chaos since his election in 2013, was flanked by high-ranking officers and his security detail at a military parade. In the middle of his speech, Maduro and others on the dais noticed a drone coming toward them and a sudden bang as it exploded. Another drone was following along and crashed into a building two blocks away, exploding as it landed on the ground.30 Maduro was protected by blast shields and rushed off the scene, but the first drone explosion injured seven members of the Venezuelan military and caused panic among those assembled. The drones were identified as DJI Matrice 600 commercial rotor drones, estimated to cost between $5,000 and $8,000 each, equipped with cameras and 2 pounds of C4 explosive.31 These drones are capable of moving very fast and covering hundreds of yards quickly as they zoom toward their target.
It is unclear why the drones exploded where and when they did, although most experts speculated that they were equipped with remote detonators. The Venezuelan military subsequently claimed that they had disrupted the drones with “special techniques and radio inhibitors,” disorienting them and making them crash before they could hit Maduro.32 It is unclear whether this was in fact the case, and whether the second drone malfunctioned or was otherwise knocked from the sky. While one anti-Maduro group implied that it had knowledge of the attack, the ultimate perpetrator of the drone attack remains unknown. The Maduro government, however, wasted little time pointing fingers at the opposition and arrested a number of those it claimed were responsible.33 The attack also inflamed international tensions, as the Venezuelan government blamed Colombia and the United States for funding and supporting “ultra-far right” figures behind the assassination attempt.34
This event also illustrated that the risk of drone assassinations is growing. The small commercial drones used in the Maduro attempt are easy to use, even by those with limited flying skills. As drones get smaller and more lethal, they will make assassination missions available to a wider range of groups. One option would be for terrorist groups to attach what are sometimes called “mini-munitions”—essentially small missiles—onto smaller drones. The US military has already found that attaching mini-munitions could make drones such as the RQ-7 Shadow capable of very precise killings of specific individuals on the battlefield.35 Some military-grade drones are specifically designed for this purpose.36 One such kamikaze drone is called the Switchblade; it was introduced by AeroVironment in 2012 and sold widely to the US military. Weighing only 6 pounds and fitting into a backpack, the Switchblade can be quickly assembled and thrown in the air for a short flight. It can also carry small munitions, enough for an explosion comparable to a grenade, and could be flown at an individual or a vehicle in an assassination attempt. While most mini-munitions and the Switchblade drone are not sold to individuals or groups other than the US military at this point, similar technology will inevitably spill out into the public domain and make assassinations easier for a larger array of non-state actors.
Another potential type of attack possible with drones involves attaching lightweight small arms, such as a pistol or a rifle, to a drone and flying it over a crowd of people to strafe them. There is some evidence that these types of drone attacks are possible, but they are harder to launch and operate than comparable attacks with IEDs. In 2015, Austin Haughwout, an eighteen-year-old college student in Connecticut, proved this was possible by developing a handgun-firing drone; the video showing its operation he uploaded to YouTube was viewed 3.7 million times within a year.37 Another one of Haughwout’s videos—called “Thanksgiving dinner”—showed a flamethrower attached to a drone roasting a turkey in his backyard. Both videos attracted interest from the Federal Aviation Administration (FAA) and local police; he and his family found themselves in court explaining why their drones did not violate FAA safety regulations.38 Although it is possible to equip drones with firearms, and this has been demonstrated in controlled conditions, such drones take more skill to use than it might appear. The recoil of the weapons tends to destabilize the drone and throw it off course, rendering them less accurate than even a terrorist might hope.39 They are also more likely to crash, and to be thwarted by countermeasures, than one which drops an explosive payload and quickly retreats.
One much-feared scenario involves terrorists loading crop-dusting drones with radioactive material or chemicals and spraying them over population centers or crowds of people. To some extent, this is a long-held dream of terrorists. Before their June 1994 attack on the Tokyo subway, the apocalyptic cult Aum Shinrikyo tested the use of remote-controlled mini-helicopters equipped with spray mechanisms to distribute sarin against their targets, but their tests failed.40 By 2003, the Bush administration was concerned that UAVs equipped with crop-dusters could be used to spread chemical or biological weapons over Western cities.41 Some analysts have argued that spraying drones would be more effective than dropping explosives and would affect a wider area.42 Al Qaeda has long had an interest in these types of attacks.43 Several of the ISIS-inspired operatives who attacked the Brussels airport in 2016 had first sought information on how to conduct a “dirty drone” attack.44 Yet the “dirty drone” scenario relies on a number of things going exactly right to work. Aside from acquiring the right kind of dangerous or radioactive materials, terrorist organizations would need to pack the material safely onto a drone, develop and test an effective dispersal system, and conduct test runs to ensure that dispersal happened in the way that was predicted.45 All of these steps would need to be conducted without attracting the attention of the security services. Even if the tests were successful, such an attack would more likely cause panic rather than many direct fatalities.46
Finally, hijacked drones are a potential threat. In a sky filled with thousands of new drones, it is possible for terrorists to find ways to seize control of drones flown by private companies or the police and redirect them to nefarious purposes. While military and some police drones have a secure connection between the drone and the operator, many commercial and hobbyist drones have an insecure communications link that could be interrupted or hacked by terrorists. If a drone hack were successful, terrorists could crash a drone or drones into a potential target or reverse engineer it for nefarious use later.47 Some have suggested that widely available commercial drones, like the quadcopters soon to be employed by Amazon, could be retrofitted to deploy explosive devices if hijacked by a terrorist group.48 While seizing control of a drone is difficult, it is possible if a hijacker is able to “spoof” the drone and send commands that the drone believes is coming from its legitimate operator. This has already happened: in 2011, Iran claims to have “spoofed” an RQ-170 Sentinel drone, sent it fake GPS signals, and convinced it to land in Iran itself, rather than Afghanistan.49 Iran then proudly displayed the captured US drone on televisions and harvested its technology for their own use. Today, these hacks are not only the province of governments. One US-based security researcher has seized control of police quadcopters with insecure lines of communication using a laptop and a cheap radio chip connected by a USB port.50 At a minimum, hijackers from terrorist groups will be able to break into drone signals and collect data and imagery from them.51 In December 2009, Iraqi insurgents managed to hack into a US Predator feed, but they did not succeed in changing its flight path.52 To do so, they used cheap, off-the-shelf software called “SkyGrabber” that is widely available online.53 Others have followed suit. In 2016, it was reported that a twenty-three-year-old hacker from Palestinian Islamic Jihad had hacked and monitored Israel’s drone feeds for at least two years.54
Some potential terrorist scenarios also rely on turning the drone into a weapon, but in a more creative way, for example directing a drone toward an engine of a manned aircraft.55 As early as 2004, German intelligence found that al Qaeda was planning on conducting these attacks, though they never got far beyond the concept stage.56 Much of the fear around this potential attack has to do with the increasing level of near-misses between drones and civilian aircraft. In the United States and many other countries, there is a clear exclusion zone around major commercial airports, and in the United States drones are not allowed to be flown above 400 feet so that they do not endanger other air traffic. Yet the enforcement of these standards has been slow and inconsistent, and drones are regularly reported flying near or inside exclusion zones and close to manned aircraft. An analysis by Bard College of 921 unexpected drone encounters reported in US airspace between December 2013 and September 2015 revealed that 90% of them occurred at over 400 feet and that a majority occurred within 5 miles of a commercial airport.57 Some of the commercial and consumer drone accidents are due to human error, but a portion are due to a malfunction with the communication signal which leads the drone to fly into congested or otherwise banned airspace. Another estimate of drone accidents between 2006 and 2016 found that 64% of incidents were caused by technical problems.58 FAA data indicates that this is a constant threat: in 2015, there were 3.5 near misses a day between drones and manned aircraft, compared with less than one a day in 2014.59 In 2016, the FAA counted 583 near misses between drones and airplanes, a threefold increase from the number of potential accidents in 2014.60 In 2017, the FAA reported a reduction in such incidents, but there were still 385 near misses with aircrafts in the United States alone, with drones accounting for more than half of reported events.61 Part of the problem is verifying whether a near miss was with a drone, a bird, or something else. While the FAA reported an estimated 6,000 sightings of drones near manned aircraft or airports between 2014 and 2018, it emphasized that it could not verify all or even most of these were actually drones.62 A more detailed analysis of near misses between December 2013 and September 2015 found that in 158 cases a drone came within 200 feet or less of a manned aircraft, and in twenty-eight cases a pilot had to rapidly maneuver to avoid a collision with a drone. Although more than half of the near misses occur near airports, some occurred further afield and above the altitude of 400 feet, where drone flights are forbidden.63
Drones are particularly dangerous around civilian aircraft because they are hard to detect on radar and because their owners are difficult to trace once an incident has been reported. In 2013, the FAA launched investigations of twenty-three incidents of illegal drone use near civilian airports or in proximity to aircraft, but in most cases the owners of the drones were never found.4 While most near misses were minor incidents, some drones came close to crashing into large passenger aircraft near major airports. In March 2013, a small private drone came within 200 feet of an Alitalia commercial jet over JFK airport, in New York.64 One year later, an American Airlines jet had a near miss with a drone in Florida.65 In 2015, a Jetblue A320 pilot reported a near miss with a drone at 6,000 feet near JFK airport, while in 2016 a Lufthansa A380 super-jumbo jet came within 200 feet of hitting a drone over LAX airport in Los Angeles.66 A prominent airline pilots association has argued that the widespread introduction of drones into domestic airspace carries could “profoundly degrade the safety of both commercial and general aviation flight operations” unless they are integrated into the FAA systems in a comprehensive way.67 Although the FAA is working on implementing regulations to integrate drones into commercial airspace in a safe way, it has not settled on a clear, uncontested set of regulations, and questions remain about tracing drones effectively in US airspace.
Britain has a similar problem: near misses with drones are increasingly common at their major airports. In 2015, the United Kingdom reported twenty-three incidents in which drones flew too close to manned aircraft in its airspace. In one instance, a drone flew within 25 meters of a Boeing 777 after it took off from Heathrow airport, and no one ever found the drone operator. An Embraer jet came within 60 feet of a drone over the Houses of Parliament.68 In 2018, a Virgin Atlantic airliner flying from Delhi to Heathrow was nearly struck by a drone on its descent: the drone came within a few meters of the aircraft, making it the closest near miss yet recorded.69
Despite these incidents, the overall odds of a collision between a manned aircraft and a drone remain low.70 Most drones fly well below the altitudes of commercial and private flights. There is also considerable debate over what would happen if a terrorist managed to make a drone collide with a commercial airliner.71 Some analysts believe that the kinetic energy of the drones would be sufficient to take an aircraft down, while others argue that a plane could easily survive a “mechanical bird strike” if the drone did not come close to a plane’s engine, or took out only one of them.72 Although there have been lab tests of this scenario, none have obviously occurred in real life, so no one knows what would happen to a plane in midflight. It also matters whether the drone would strike the aircraft at a vulnerable point in its flight, such as during take-off or landing. The conventional wisdom is that interruptions of the aircraft’s operation at those points are far more perilous than during cruising, but the size, speed, and angle of strike for the drone aircraft would still influence the severity of the damage caused.
Part of the reason why what might happen with a drone hitting an aircraft remains unclear is that near misses—or at least reports of them—are now common, but actual strikes of drones into manned aircraft are extremely rare. A widely noted report of a drone striking a British Airways jet near Heathrow airport in April 2016 turned out to be a false alarm.73 There are only two verified accounts of a drone striking an aircraft and neither resulted in the catastrophic loss of the aircraft. In Canada, a small drone collided with a Skyjet plane making an approach to Quebec City in October 2017, but the plane landed safely after the pilots heard a loud bang.74 In 2017, a drone was flown into a Black Hawk Army helicopter over New York Harbor, but caused little damage.75 Although accidents are always possible, it is perhaps even more difficult for terrorists to deliberately strike an aircraft flying at 500 feet per minute with a drone unless the navigation, accuracy, and speed of small commercial and hobbyist models improve significantly. DJI Technology, the largest commercial drone manufacturer, has argued that these attacks are unlikely, because it would be “like trying to hit a bullet with a bullet” and would require “an unprecedented act of marksmanship.”76
It would be far easier to interrupt commercial flight operations with drones, while not actually striking an aircraft. Over the last few years, drone sightings near airports have become commonplace and have shut down airports across the world. Some of the biggest airports in the world, including Frankfurt, Dubai, Singapore, London Heathrow, and Newark have been shut down temporarily due to drone overflights. In many cases, a pilot or other airport official will report seeing something like a drone near the airport, but verifying that it actually was a drone is difficult.77 This was illustrated by the sighting of one or more drones near London’s Gatwick Airport over Christmas 2018. The airport was shut for three days with multiple reports of drones over landing strips, and over one hundred flights were canceled over the busy holiday period, affecting approximately 140,000 passengers. But the event remains a mystery: despite multiple credible reports of sightings, no video has emerged to confirm that it was a drone, and no drone has been found despite an extensive search of the surrounding area. One Sussex police officer even remarked that it was “always a possibility that there may not have been any genuine drone activity in the first place.”878
The mystery deepened when the subsequent investigation revealed no obvious culprit. Two people were arrested but were quickly released when British police admitted that they were no longer suspects.79 The police reportedly suspect that this may have been an inside job by someone who once worked at Gatwick because the perpetrator knew the airport layout and the blind spots—for example, behind buildings—where counter-drone technology would not work.80 Part of the problem the police faced was finding the drone: signal-jamming counter-drone technology is risky to use near an airport, and some more crude ways of getting the drone—for example, shooting it down—would be prohibited near airports. Although the military was called in to find the drone, they had no more luck than the police. In the end, the motive was equally obscure: police were uncertain whether the event was an act of terrorism, a protest of some kind, or criminal mischief.81 Ultimately, it did not matter, because the Gatwick incident revealed that drone disruptions like this could be accomplished without the perpetrators getting caught provided that the attack was well planned. The attack was also cost-efficient: for the price of a single drone and related equipment, the attacker cost Gatwick airport £1.4 million and cost the airlines £64.5 million.82 Although London Gatwick and other major airports have rushed to buy counter-drone technology that will prevent such attacks in the future, the success of the Gatwick airport disruption has surely been noticed by terrorist groups.83
How likely are these scenarios? Any calculation of risk from terrorist drones is partially determined by the number of drones available. As drones wind up in the hands of hundreds of thousands of new people each year, the risk that someone will find a way to misuse them and to harm others naturally increases, as it does with many other types of technology. One reason why government officials have been so concerned is that many had vastly underestimated just how many people would acquire drones for commercial and personal uses. The rapid proliferation of cheap, small drones—particularly the increasingly sophisticated commercial off-the-shelf drones, available for only a few hundred dollars from Amazon and other retailers—means that it is inevitable that the technology will wind up in the wrong hands and be put to criminal purposes. This can already be seen in the use of drones for delivering contraband into prisons. In the United States alone, there were twelve attempts to drop mobile phones, drugs, and pornography into prisons between 2012 and 2017.84 It is likely that this is a substantial underestimate and that the number today is far higher. For example, Mexican drug gangs have been using drones to move drugs and contraband and to monitor smuggling routes.85 They have also weaponized small commercial drones by equipping them with explosives to use against their rivals in the drug war.86
For terrorists, the barriers to entry are generally higher but they are not insurmountable. Many attempted drone attacks will fail or be thwarted, but if enough attempts are made, at some point the odds are that an attack of some kind will succeed. It is also likely that terrorist interest in drones—and hence attempted attacks—will increase as their availability and capabilities increase. There are at least five reasons why terrorist organizations might want to turn to drones over the next decade or more. First, for organizations like al Qaeda and ISIS, who struggle to get their operatives into the United States, it is inefficient to use suicide attacks and waste those assets; ideally, a reusable means of attack, like a drone, is better.87 Terrorist organizations would prefer not to have any operatives near an attack, but would instead prefer to program the drone to attack on a predetermined flight path, as happened with the Maduro assassination attempt. This is now possible through freely available open source software such as ArduPilot.88 Second, if small drones could reach their target quickly, it would dramatically shorten the time that law enforcement authorities have to react. One senior British government official noted that “once you start to think of a large hobbyist drone carrying a standard payload, most of the traditional defenses against terrorist penetration of high-value places become irrelevant.”89 Third, the lack of adequate regulation and control of drones also provides an opportunity for terrorists. An off-the-shelf hobbyist drone can be bought without attracting the attention of most regulatory officials or law enforcement. The sheer volume of drones available on the commercial market, and the inadequate records of who has purchased a drone, will mean that attribution for an attack will be difficult. The time lost in finding who really owns a drone might give terrorist operatives the opportunity to escape detection and capture. Fourth, drones also allow for forward planning. The increasingly sophisticated cameras on drones allow for terrorist operatives to scout out future attack locations and conduct dry runs to ensure that their attacks work as planned. Fifth, drones are adaptable: they can be amended with new cameras or payload, and new technology, like 3-D printing and Openware, provides an opportunity to tailor the drone’s operations for an attack. As off-the-shelf drones become smaller, more portable, and more capable, they can be adapted more quickly to a wide range of purposes, including terrorist attacks. This is particularly the case given the wide availability of information on adapting drones available through online forums like DIY Drone. Similarly, the integration between small commercial and hobbyist drones and widely available consumer electronics—WiFi, GoPro cameras, and even iPads and mobile phones—will lower the technological barriers to effective drone use for terrorists and enhance adaptability.90 The more that drones adopt “plug and play” technology, the easier terrorist drone attacks will become, especially for novices or lone wolves.
There is no doubt that evidence of terrorist interest in drones is growing. One unconfirmed report suggested that authorities in the United States, Germany, Spain, and Egypt foiled at least six potential terrorist attacks with drones between 2011 and 2015.91 There is no estimate of the number of foiled attacks more recently, but given the spread of the technology it is likely to be higher. There is evidence that drones are in the hands of a number of terrorist organizations, including al Qaeda, the Taliban, Fuerzas Armadas Revolucionaria de Colombia (FARC), Hezbollah, Hamas, and the Islamic State, although mostly on foreign battlefields.92 Evidence of actual attacks, or even scouting for attack sites, is fragmentary. There have been reports of unauthorized use of drones in sensitive sites in the United States, the United Kingdom, and France, but it is unclear how many of these were genuine attempts to scout locations for future attacks. The most famous drone accident in the United States looked for a brief moment like an attack. A drunken US government employee accidentally flew his drone onto the lawn of the White House, and then fell asleep.93 The incident was comical, but the implication that a terrorist drone could penetrate the protected airspace above the White House was widely noted. In the United Kingdom, the Metropolitan Police has acknowledged twenty suspicious drone incidents in and around London, though most were violations of airspace or criminal activity.94 In France, unidentified drones have been flown over a number of sensitive sites, including the Eiffel Tower, Place de la Concorde, and Elysee Palace, and over multiple nuclear power stations.95 An unidentified drone even allegedly struck the Sydney Opera House in October 2015.96 Yet despite these near misses, no terrorist organization has successfully used a drone for an attack against a fixed location or major grouping of people, such as a sporting event, to date.
This situation—that drones are widely available, interest is high among terrorist organizations, but attacks have not yet occurred—is due in part to the formidable but underappreciated technological obstacles to launching these attacks. Put simply, imagining a “terrorist drone” attack scenario is easy, but conducting that attack is harder than it looks. Even if drones are readily available and terrorist interest in them is growing, there are practical obstacles that must still be overcome. First, drones have limits in range and endurance that make it hard to plan an attack at a great distance. Most commercial off-the-shelf drones are able to sustain a datalink of 1–10 km from a ground station, with many of the less regulated models on the lower end of that range.97 What this means is that with current drone technology terrorists would need to be relatively close to the site of the attack for it to work. These communications and data links are often unsecured, which leaves the plot at risk of being detected by the police or security services. Although higher end commercial drones have an endurance limit of several hours, many hobbyist drones have only an endurance of under an hour.98 While this can be modified, doing so will affect payload and limit what kinds of explosives might be carried. Given these technical limitations, most terrorists would need to get somewhat close to their proposed target and assume some personal risks that they will be intercepted before the attack proceeds. Especially in cities, where their activities may already be monitored by the police and security services, this presents some real dangers. This is obviously not insurmountable: the perpetrators of the Maduro and Gatwick events got away with it, but clearly not everyone else will. Finally, some contemporary drones are slow and could be seen on approach before an attack. Most hobbyist rotary wing drones are not optimized for speed and can travel at a maximum of 15–20 meters per second (49–65 feet per second). This leaves them susceptible to anti-drone technology and other cruder ways of stopping a drone in mid-flight if they are launched a decent distance from the target. Until drones get faster or move in swarms, current generation drone attacks are likely to be spotted or even heard, given their whirring in-flight noises, before they can reach their target.
Arming drones also presents its own challenges. Many of the alarming scenarios of terrorist drones tend to underplay the real difficulties involved in deploying dangerous materials—for example, radiological material, chemical weapons, and others—onto mobile flying machines. The chemicals and radioactive materials are under varying degrees of scrutiny and control; purchasing or otherwise obtaining them can draw the attention of the police or the security services. Many of these payloads are unstable and carry real risks for those installing and deploying them; some, such as radioactive materials, require careful storage and handling that is difficult to provide outside a lab. Even simple explosives are difficult to handle and can backfire, leading to the death of the planners themselves. Moreover, these heavier, more unstable explosive payloads can reduce the stability and the payload capacity of a drone. One estimate found that the popular DJI Phantom 3 model lost 14.4 minutes of flying time for each 1 kg of payload it assumed.99 Large payloads of explosives can destabilize the drone and cause it to crash; destabilization and crashes are even more likely when the drone is rigged to firearms and must remain stable during recoil.
A terrorist contemplating using a drone for an attack would have to weigh the advantages conferred by the technology against the technological obstacles to their use. It is one thing to simulate a terrorist attack with a drone in a laboratory or in controlled conditions, but it is quite another to conduct it in real life. Much of the planning for a terrorist drone attack—for example, buying the necessary components or conducting dry runs against fixed targets—may attract the attention of the police and security services, especially in inhospitable environments like major cities in the developed world. In other words, “terrorist drone” attacks are possible, but they are harder than some dire warnings suggest. It would obviously be foolish to declare that terrorist drone attacks will never happen given the explosive growth of drone technology worldwide and the evident terrorist interest in the technology. But equally taking the nightmare scenarios at face value and hyping the threat flowing from them is wrongheaded. The risks posed with current drone technology are real but not insurmountable, especially given the growing attention and investment among governments, private companies, and others in counter-drone technology. Given these risks and barriers, successful terrorist drone attacks will generally require some basic technical knowledge and organizational capacity by the perpetrator. Among non-state actors, these characteristics are more likely to be found in disciplined, well-resourced rebel groups in today’s theaters of war than among resource-constrained terrorist cells or angry “lone wolves” in the developed world.
Much of what has been written on terrorist drones conflates their use by rebel groups in theaters of war in Central Asia, the Middle East, and elsewhere with their potential use by terrorists in cities like New York, London, and Paris. Government officials and experts often point to evidence of the use of drones by rebel groups fighting wars abroad as showing why terrorist drone attacks on targets in the developed world are likely today. Yet the context surrounding these rebel drones is very different. The terrorist organizations that have been most successful in their use of drones—for example, Hezbollah in Lebanon and the Islamic State in Syria—have deployed them locally in their capacity as rebels engaged in a violent struggle against a government. In these wars, the targets are generally less well defended and the risks of being caught are lower than they are in well-monitored cities. War is a naturally permissive environment that enables experimentation with and testing of the means of violence; that testing is further possible because rebel groups remain in regular communication with military command and supply lines that put the technology in their hands. Many of the difficulties present in launching a terrorist attack in the developed world are simply not present in states experiencing an active armed conflict. For these reasons, it is far easier to fly a rebel drone than a terrorist drone, and it is far more likely that drone deployment will help level the playing field for rebels against their government opponents.
The emergence of “rebel drones” signals a significant change in warfare in three ways. First, for many years, militaries from the developed world have made achieving air superiority their first goal when undertaking any military operation. For example, the US Air Force insists on air superiority in counterinsurgency, peacekeeping, and other missions and destroys any enemy aircraft before they can even take to the skies. This gives the United States complete freedom to attack its enemies with air power and enables ground forces to undertake maneuvers that would otherwise be too risky. But with drones, rebel groups and insurgents can now take to the skies themselves and challenge the assumption of air superiority that underpins most of their opponent’s military strategy. Most commercial drones are small enough and fly at such low altitudes that they evade detection by normal radar. Second, most asymmetrical conflicts do not distribute vulnerability evenly but instead generally leave rebel groups or insurgents at greater risk of death and injury than troops from the developed world. This is particularly the case with US forces that have high standards of force protection. While rebel drones do not level the playing field completely, they shift vulnerability toward soldiers on the ground by leaving them subject to attack from the air, sometimes for the first time in decades. Third, these rebel drones have a psychological impact on their opponents by illustrating their vulnerability to an attack which may come without warning. The psychological aspect of drones—that they can come anytime, unexpectedly, and cause damage—is as important as their immediate tactical value, if not more so.
The earliest example of a rebel use of a drone provides an illustration of how a permissive environment and organizational infrastructure is an essential foundation for rebel drone use. A Colombian army unit stumbled across nine remote-controlled planes belonging to the FARC rebel group in the remote jungles of Colombia in August 2002.100 The purpose of the remote-controlled planes remains unknown, but army officials suggested that they were intended to launch IEDs against government targets.101 This example is often cited as the beginning of the terrorist drone threat, but in fact the FARC was hardly a normal rebel group. At the time, it was at its most powerful, controlling vast swaths of territory inside the country and managing a significant portion of the drug trade. It was, in other words, not a ragtag group on the margins of the society, but rather a powerful army with substantial membership, funding, and infrastructure.102 It is not surprising that such a well-equipped and capable armed group would be among the first to truly experiment with launching drone-based attacks.
The same could be said of Hamas, the terrorist organization that has effectively controlled the Gaza strip since the Israeli withdrawal in 2005. Its military wing, the al-Qassam Brigades, has a small number of drones and has attempted to penetrate Israeli airspace and to launch attacks on Israeli targets with them.103 In November 2012, Israeli military officials discovered and destroyed a drone workshop run by Hamas in Khan Yunis in the southern Gaza strip.104 In 2013, Hamas operatives attempted to pack a drone full of explosives and send it into Israel, but the plot was disrupted.105 In July 2014, Israel shot down a Hamas drone with a Patriot Missile, although Hamas later claimed that two additional drones made it through undetected. Hamas has clearly exploited drones for their psychological impact, even flying a drone over a military parade in December 2014.106 Hamas claims to have a wide range of surveillance and attack platforms and releases grainy images of its drones after successful flights. They have continued to send small drones from the Hamas-controlled Gaza Strip over Israeli territory. For Hamas, drones are remarkably cost-effective against the more powerful and well-resourced Israeli government. For only a few hundred dollars spent on a commercial drone, they can force Israel to scramble its jets and spend thousands or more to shoot them down. Unsurprisingly, Israel has retaliated by regularly striking their drone factories and allegedly by killing a Tunisian man who led a double life as a Hamas drone engineer.107
The armed group that has been most successful at using drones is Hezbollah, one of the most organized and disciplined non-state actors in the Middle East. Hezbollah is a complex organization, operating more like a powerful political and military player in Lebanon, but also employing terrorism against Israel and other targets in the Middle East. By some estimates, Hezbollah now has more than 200 platforms for reconnaissance and combat missions, with a significant portion of its drone fleet provided by Iran.108 In September 2013, Hezbollah became the first non-state actor to launch its own successful drone strike, dropping either explosive warheads or air to ground rockets on the Jabhat al-Nusra forces in an attack that killed twenty-three al-Nusra fighters.109 It is also the only non-state actor with its own dedicated drone airfield. In April 2015, it was reported to have a constructed an airfield and sizeable drone fleet in the Beka’a Valley in Lebanon.110 By August 2016, Hezbollah was dropping Chinese-made cluster bombs from drones on buildings and cars occupied by Syrian rebel forces.111 In doing so, Hezbollah became the first rebel group to do what the United States has long done: use drones to eliminate its enemies through targeted killings. It has continued to launch drone strikes against its enemies in the Syrian civil war, but its experience with drones predates that by some time.
The first evidence of Hezbollah mastering a drone came shortly after the onset of the second intifada against Israel. In December 2003, a Hezbollah cell was caught by Israeli security officials working with the al Asqa Martyrs Brigade, an armed wing of Fatah, to plan an IED drone attack on Jewish settlements in Gaza.112 After that attack was foiled, Hezbollah shifted toward deploying drones as an almost symbolic show of force over Israeli territory. In November 2004, it launched a Mirsad-1 drone for a twenty-minute reconnaissance mission over northern Lebanon, though the drone crashed shortly thereafter. Hezbollah’s leader, Hassan Nasrallah, boasted that the group could strike “anywhere, deep, deep” inside Israel and could deploy as much as 200 kg of explosives.113 While this may have been an empty boast, Hezbollah successfully continued its drone program with a successful drone launch in April 2005, when another Mirsad-1 drone hovered over the Israeli city of Acre and returned to Lebanon before it could be destroyed.
During the Lebanon War in 2006, Hezbollah launched several drones against Israeli targets and even rammed an explosive-packed drone into an Israeli warship, causing a small fire.114 By August 2006, Hezbollah’s capabilities had improved and it was able to launch three small Ababil drones—packed with 40–50 kg of explosives in warheads—against targets in Israel.115 While these were shot down, they illustrated that small drones could evade detection by Israel’s sophisticated air defense systems.116 These attacks were negligible from a military vantage point, as Hezbollah’s drones were shot down by manned aircraft and paled in comparison to Israel’s sophisticated drones. Yet they proved to be a powerful signal of capacity and source of propaganda for Hezbollah. Matthew Levitt of the Washington Institute for Near East Policy has remarked that, “they love being able to say ‘Israel is infiltrating our airspace, so we’ll infiltrate theirs, drone for drone.’ ”117
Figure 5.1 Pieces of a fixed wing Hezbollah UAV downed by the Israeli air force in 2006.
Although Hezbollah may love claiming credit for its drones, it is not clear whether it would have been capable of using drones in the absence of external sponsorship. The Ababil drones—named for a mythical race of birds which dropped stones on an army invading Mecca, according to the Quran—have been exported by Iran widely to both Hezbollah and Hamas, both determined enemies of Israel.118 Iran has also sold the Mirsad drone—an updated model of a reconnaissance drone flown since the Iran-Iraq war in the 1980s—to Hezbollah.119 There are unconfirmed reports that Hezbollah has Shahed-129 drones, which are like smaller, less-capable versions of the Predator.120 Hezbollah has insisted that these drones were not provided by Iran and were developed by its own engineers, but the available photographic evidence shows a clear resemblance between Iranian models and their Hezbollah counterparts. There are unconfirmed reports from Russian and Israeli sources that Iran sent eight Mirsad drones to Hezbollah and even trained thirty Hezbollah operatives near Isfahan to fly the aircraft.121 Expertise from Iran is also flowing to Hezbollah’s engineers, allowing them to build a stronger capability for attacks against Israel. This close relationship between Hezbollah and Iran led Milton Hoenig to conclude that “drone launches by Hezbollah into Israel are planned and carried out to meet the political agenda of Iran, while shielding Iran’s involvement and allowing a measure of deniability.”122
The relationship between client and sponsor may be more complicated than this depiction suggests. While Hezbollah may do the bidding of its Iranian sponsor in some instances, in others it is clear that Hezbollah is using its drone fleet for its own priorities, specifically deploying them in an almost theatrical way to show off its capacity to threaten Israel. None of the drones deployed by Hezbollah are comparable in capacity to those flown by Israel; in fact, many are somewhat low tech, flying at such a low speed and elevation that they are hard to detect on radar.123 These relatively simple drones have been detected by Israel’s sophisticated Iron Dome defense system and shot down. While their military utility is limited, they have some value in scaring Israeli civilians and reminding them they remain as vulnerable to Hezbollah’s drones as Lebanese and Palestinian civilians are to Israel’s drones. In other words, Hezbollah’s drone incursions are more about psychological warfare than military strategy. They serve as a symbol of Hezbollah’s persistence and a signal of its capabilities, even if those capabilities are vastly outmatched by its opponent.
This approach was clearly apparent in Hezbollah’s second wave of drone flights in 2012. After a six-year moratorium, Hezbollah launched an Iranian drone near the Israeli nuclear reactors in Dimona, approximately 35 miles within Israel’s territory across the Negev desert.124 It was shot down by Israeli aircraft, but not before it was rumored to have taken pictures of sensitive nuclear sites.125 Within days, Hezbollah was exploiting the operation for propaganda purposes. Hassan Nasrallah, the leader of Hezbollah, noted that Israel frequently sent drones into Lebanon’s airspace and that “it is our right to send other drones whenever we want.” He further warned, “It was not the first time and it will not be the last.”126 As if to prove his point, a further drone incursion was reported in April 2013.127 Neither of these drones posed a military threat, but both had the effect of allowing Hezbollah to claim a propaganda victory and to deliver a shot across the bow of its enemy Israel. This psychological impact is only possible because actors like Hamas and Hezbollah are connected to the funding, infrastructure, and supply lines needed to make these drone operations possible.
It is often said that experience is a harsh teacher.128 The experience of war has proven valuable, however, for rebel groups who have begun to experiment with drones. While the use of drones by non-state actors in the mid-2000s revolved around sporadic tactical attacks and signaling, rebel groups have more recently shifted to using drones as a tactic in a range of conflicts, including Afghanistan, Libya, Syria, and Ukraine. For some rebel groups, this meant stealing or borrowing a Western military grade drone for their own battlefield use. In May 2012, a NATO raid in Helmand province in Afghanistan discovered that the Taliban had captured a small drone, possibly modeled on NATO’s own Desert Hawk.129 In Libya, government and militia forces have deployed a wide array of commercial and hobbyist drones for battlefield reconnaissance, some of which came from the West. In 2011, Libyan anti-government forces even managed to acquire a $120,000 quadcopter from the Canadian firm Aeryon Labs for battlefield reconnaissance.130 But drone use—by the United States, Turkey, United Arab Emirates, and others—escalated throughout Libya’s civil war, and their practice was matched by non-state actors. By 2019, Libyan National Army (LNA) forces were being supported by armed Chinese Wing Loong drones operated either on their own or by their allies, the United Arab Emirates.131 In June 2019, LNA forces also destroyed a Turkish drone that they claimed was targeting their positions.132
The rebel groups that manage most consistently to deploy their own drones for strategic effect have a strong organizational base, a steady flow of funding, and one or more external sponsors. These groups use battlefields to hone their skills, to experiment, and to learn from their failures in a complex, violent environment. But beyond just using battlefields as incubators for drone expertise and development, rebel groups fighting in Ukraine, Iraq, and Syria have also used the technology to level the playing field against their more powerful opponents, turning asymmetrical wars into wars where governments and rebels are in a race to deploy drone technology in more innovative ways.133 In the case of the Islamic State, it also led them to expand their goals and made them willing to undertake attacks against US forces that would have been unthinkable before.
When Russian forces invaded Crimea and later eastern Ukraine in 2014, combatants on both sides turned to drones for a wide array of reconnaissance activities. Both sides were seeking battlefield awareness: in the fog of a proxy war, with pro- and anti-government rebel groups operating without uniforms in an environment with civilians present, getting information on the location of the enemy becomes paramount. Pro-Russian separatist forces turned to Moscow to provide them with drones. A number of Russian-made drones, including the Orlan-10, Eleron-3SV, Granat-1, and ZALA drones, have been shot down in Ukraine.134 One estimate suggests Russia has deployed as many as sixteen drone prototypes there.135 Most of these drones are broadly similar to the MQ-11 Raven drone and are highly effective reconnaissance platforms. They allow pro-Russian rebels to identify targets, such as pro-government rebel positions, and direct artillery fire or other types of bombardment.136 A senior US military official observing the conflict found that this tactic of drone spotting and bombardment had been brutally effective, with 85% of Ukrainian casualties coming from rocket and cannon fire.137 Pro-Russian separatist rebels had also purchased commercial off-the-shelf drones and found ways to strap explosives to them and even drop grenades with them.138 According to one analyst, the successful exploitation of drones poses a real risk that the conflict would be determined by “technical overmatch.”139 Knowing that Moscow’s supply of drones would reveal its hand behind the conflict, pro-Russian rebels have also deployed GPS spoofing technology and signal jammers to block Ukrainian and Organization for Security and Co-operation in Europe (OSCE)-funded drones from collecting evidence of Russia’s direct involvement in the conflict.140
Ukrainian military forces have been at a substantial strategic disadvantage because the government had underinvested in drones and had relatively few ready for the conflict.141 To compensate, the Ukrainian government turned to commercial, off-the-shelf drones as a way of evening the score.142 It has mostly employed quadcopters with commercial cameras for reconnaissance of the battlefield and identification of potential targets. The Ukrainian military also created a special military unit, the Aerorozvidka, to train pilots in the effective use of commercial drones.143 To more quickly make up the “drone gap,” Ukrainian pro-government forces have turned to crowdfunding to get drones to the battlefield. For example, one organization, the People’s Project, operated its own version of a Kickstarter fundraiser to buy drones.144 One US organization, the Chicago Automaidan, has also been purchasing and retrofitting DJI Phantom drones for military use and sending them to pro-government forces.145 Other more locally homemade drones have been also deployed by Ukrainian military forces. The United States also provided some RQ-11 Raven drones to Ukraine’s military, but these have proven less effective than hoped.146 Although they remain outmatched by pro-Russian forces in tactical terms, the Ukrainian rebels have skillfully used drones for propaganda purposes and even revealed the existence of a long-denied Russian encampment in eastern Ukraine.147 They have also periodically succeeded in dropping grenades from their drones against enemy positions and fixed targets.148 Although the introduction of drones has not leveled the playing field between pro-Russian separatist and Ukrainian government forces, it has reduced the gap in capabilities between them.
Similarly, the regional war in Syria has also proven to be a fertile testing ground for rebel groups to try their hand at flying drones. In many respects, this brutal conflict, stretched across the borders of Iraq and Syria, has been a laboratory of rebel drone innovation, with actors on all sides deploying drones for increasingly complex tasks. Within this battlespace, multiple armed actors—Kurdish Peshmerga, Jabhat al-Nusra, Islamic State, Hezbollah, and an array of militias allied to the Syrian government and to Iran—have employed drones on the battlefield. The frequency and sophistication of drone use has varied, but together these rebel organizations have shown that drones can redress some inequalities on the battlefield, broadcast their cause to the wider world, and in some cases expand the ambitions of their users.
One way that drones can redress inequalities on the battlefield is by enhancing reconnaissance of the enemy’s positions. This is a crucial development because most rebel groups face chronic manpower shortages and are reluctant to lose personnel when they are overmatched by their opponent. Without drone technology, many rebel groups lacked the ability to look around a corner and see the enemy’s position from the air, a skill that the United States and other developed countries have had for years. But this is now changing, giving them a point of view that they never had before. In Syria, pro-government militias were using DJI Phantom quadcopters for battlefield reconnaissance as early as 2013.149 Iranian-backed Syrian militias, such as Saraya al-Khorasani, have also used simple hobbyist drones to provide reconnaissance on Islamic State positions.150 The Kurdish Peshmerga have used small, fixed-wing drones to assist their operations and even worked with a US entrepreneur to acquire an LA-300 drone that boosts their capacity to see the battlefield.151 They have supplemented these with drone imagery provided by the US military, although they complain that it often arrives too late to be useful.152 In May 2015, the Kurdish Peshmerga shot down an Islamic State drone that was monitoring their position.153
All sides in the Syrian conflict have turned to drones—including commercial and hobbyist drones—as a way of advertising their successes on the battlefield and drawing support to their cause. The war in Syria is one of the first in which multiple combatants are directly filming and producing propaganda videos for upload to YouTube and other file-sharing sites. These videos are often striking in their visceral detail of destroyed buildings, combat operations, and even deaths. They are producing a new kind of spectator war, in which battlefield imagery is broken down into vignettes of video-game-like destruction for propaganda purposes. For this reason, drone imagery is now often slickly edited and produced to maximize its propaganda value. Russia Works, an organization closely linked with Russian state television, has used DJI Phantoms to produce videos of Russian soldiers and tanks in the middle of the fight in Syria to sell the war to the Russian public.154 Some Russian-made videos emphasize the careful, humane use of force that Russian forces allegedly use while others show drones sweeping over the destruction of cities like Palmyra to highlight the barbaric behavior of the Islamic State.
Non-state actors have been equally adept at making propaganda videos with drones. In 2014, al-Qaeda affiliated Jabhat al-Nusra released a well-crafted propaganda video called “Breaking the Siege,” which showed the rough outlines of their operations to rescue prisoners held by the Islamic State.155 By 2015, al-Nusra had produced slick top-down footage of its operations in Aleppo in that year, and even produced a video showing the execution of a vehicle-borne suicide attack.156 Both appeared to be taken with a standard hobbyist quadcopter and nothing more sophisticated than a GoPro camera. By 2016, they were producing widescreen shots of battles which allowed the viewer to track the movements of vehicles and the impact of cruise missiles.157 The growing technological and artistic sophistication of their videos shows that al-Nusra was aware of the need to rise above the din of the other videos of the Syrian war available online and to match those made by others, including the Islamic State. The Islamic State began releasing propaganda videos with “The Clanging of the Swords, Part 4,” a May 2014 video that played like a Hollywood trailer and showed an Islamic State drone over Fallujah in Iraq.158 It gradually developed a sophisticated infrastructure to build drones and to show off their battlefield exploits. The Islamic State has released videos of its fighters controlling drones and directing ground forces and even suicide bombers, while carefully editing the videos for propaganda purposes.159
While drones were initially used more for propaganda than tactical advantage, this has now changed. Today, there are numerous reports of the Islamic State using small, commercial, off-the-shelf drones for battlefield reconnaissance purposes and support of artillery fire.160 Of all of the rebel groups fighting in Syria, the Islamic State has deployed the greatest variety of rotary and fixed-wing drones and has produced a large number of videos interspersing drone footage with other battlefield imagery. Banned from receiving most military-style drones, the Islamic State has concentrated on acquiring Chinese-made commercial drones (such as DJI Phantom FC40) and also X-UAV and Skywalker X8FPV fixed-wing models161 and have also worked with external suppliers in Europe to get their hands on commercial drones.162 The United States has been destroying Islamic State drones more frequently as they approached the location of US troops and allies, often by jamming their signal or just shooting them down. The United States has also begun killing IS drone-makers and targeting trucks with IS drones inside.163
The Islamic State has also been experimenting with attaching improvised explosives to drones. Since they acquired surveillance drones in mid-2014, experts have worried that it was just a matter of time before they could “jury-rig surveillance drones into flying IEDs.”164 By December 2015, these fears were realized: the Islamic State packed a drone full of explosives for an attack, but it was shot down by Kurdish forces before it could be successful.165 These bomb-laden drones were little more than cheap quadcopters, but they nevertheless represent a growing threat.166 In August 2016, the US Army identified small quadcopters packed with explosives as “the greatest challenge for Army forces” when it comes to air defense.167 In October 2016, these fears were vindicated. Two Kurdish soldiers were killed and two French special forces operators were wounded when a booby-trapped drone operated by ISIS exploded near Mosul in Iraq.168 By early 2017, it was reported that the Islamic State had killed about a dozen soldiers and wounded more than fifty in approximately eighty missions in which bombs were dropped from quadcopters.169
Explosive-laden kamikaze drones have also been used by Houthi rebels in Yemen to even the playing field with Saudi Arabia. In Yemen’s civil war, Saudi Arabia and the United Arab Emirates have been fighting alongside Yemen’s military to restore its government after an overthrow attempt by Houthi rebels. The Houthi rebels, clearly outmatched by Saudi-backed forces and attacked from the air, resorted to Iranian-made drones to destroy missile defense systems provided by Saudi Arabia. 170 Over time, they expanded their ambitions with drones and brought the war home to their enemies. In July 2018, Houthi rebels claimed to have attacked an oil refinery in Saudi Arabia, and they continue to penetrate Saudi airspace to threaten its government.171 They also claimed an attack on Abu Dhabi airport in the United Arab Emirates.172 By January 2019, they had launched an Iranian-made Ababil drone equipped with a bomb to kill Yemeni military officials, including the head of the military intelligence division.173
The next step in the evolution of rebel drone attacks will involve swarming fixed locations with small, disposable drones packed with explosives. In January 2018, Russian forces in Syria reported that two of their bases there fell under simultaneous attack by DIY drones packed with explosives.174 Thirteen drones, all fixed wing, made of plywood and with a crude engine, swarmed Russian air bases and had to be disabled or shot down. Russia has blamed Syrian rebel forces for the attack, but also pointed a finger at Turkey for backing the rebels responsible for the attack.175 These swarming attacks will continue in the future. In February 2018, Houthi forces claimed to have been experimenting with drone swarm attacks to disable air defense systems, although it is unclear whether these attacks have been successful.176 Notably, in both cases, it was the external sponsorship of a government—Turkey and Iran—which allowed the rebel groups to target their enemies in such a creative way.
We are now living in a world where terrorists at home and rebel groups fighting on distant battlefields have as much access to small drones as consumers in the developed world. There is no natural limit to their ingenuity: with these tools in their hands, they will find creative ways to learn about their enemies and target their weaknesses. With drones diffusing rapidly across the world, the greatest advantages will flow to those non-state actors with the organizational infrastructure and budget to acquire and deploy drones effectively. Just as is the case with states, there is a limited first-mover advantage: those actors, like Hezbollah, who master drone use earliest will be more capable of using them to strategic ends than those who come to them later. Similarly, those rebel organizations with a generous external sponsor (like Russia, Turkey, or Iran) will have better access to sophisticated drones, and thus better results, than those who can only purchase them from commercial outlets.
At this point, drones are not a total game-changer for most non-state actors. They will have the greatest impact in leveling the playing field in conflicts where the asymmetry of power is relatively slight between the major players. In these cases, the use of even small commercial or hobbyist drones for reconnaissance will cut into the advantages of more powerful fighters. This dynamic could be seen in Nigeria, where the Islamist group Boko Haram acquired drones that matched or even bettered those held by the government, which allowed them to conduct more sophisticated attacks against the Nigerian military and civilian targets and expand their geographic reach across the countryside.177 They may also affect the duration of the conflict by allowing weaker rebel groups to protect their forces, harass their enemies, and survive longer even when they are obviously overmatched. Retrofitting commercial drones with explosives may provide a short-term tactical advantage to rebel groups in evenly matched wars, but it is unlikely to turn the tide in one direction. In other cases where that power asymmetry is stark—for example, Hamas attacks on Israel—drones will be of less importance. In these cases, drones may enhance the vulnerability of a powerful government by directly threatening its civilians with sporadic attacks and have an array of psychological and symbolic consequences. Against a well-armed, well-droned opponent like Israel, drones in the hands of a group like Hezbollah will not tilt the strategic balance, but may allow them to compete more effectively and contemplate doing things that they would not have done otherwise.
This suggests that the ability of non-state actors to use drones is highly context-dependent. Much of the discussion of the threat emanating from terrorist actors possessing drones conflates their use on open battlefields with their use against well-policed targets in the developed world. These are environments with different opportunities and constraints: a “rebel drone” is not necessarily a “terrorist drone.” That the Islamic State can conduct battlefield reconnaissance in Syria does not imply that it would be equally capable of doing so in Paris or London. That Hezbollah can launch drone strikes in Syria does not imply that it could do so against Western targets or even well-defended areas of Israel. With current drone technology, the ability of most terrorist organizations to use drones in non-permissive environments is likely to be limited. Even organizations like the Islamic State which fight on a battlefield and conduct attacks in Europe and the United States will find barriers to transferring their expertise and technology between these theaters of activity. While it would be a mistake to rule a terrorist drone attack out, or to dismiss the threat entirely, it is equally a mistake to give in to some of the hype surrounding this threat and assume that a terrorist drone attack on a Western city is imminent. Instead, we should expect that swarming attacks with cheap drones will become regular events in asymmetrical conflicts around the world but that terrorist drone attacks will be rare events in the developed world.
In the long run, perhaps the most important factor influencing the likelihood of a successful terrorist drone attack is the development of counter-drone technology. As more terrorist organizations are developing drones, the US government and private industry have turned to new technologies to block, track, and disrupt future attacks. In July 2016, the US Department of Defense requested an additional $20 million from Congress to address the new drone threat from the Islamic State.178 The US Treasury is now sanctioning foreign companies involved in transferring drone materials to terrorist organizations like Hezbollah.179 The Department of Homeland Security (DHS) is actively investigating counter-drone technology and even reportedly cooperated with the Department of Defense and the New York Police Department on a secret test of microwave-based counter-drone technology on New Year’s Eve in 2015.180 The FAA’s efforts to create a master drone registry, with traceable serial numbers, will go some way toward undermining one of the chief advantages of drones for terrorist groups: the lack of attribution to a person responsible for the conduct of that drone. If drones have serial numbers equivalent to Vehicle Identification Numbers (VINs) and are registered with the government, it will be harder to sell them on the black market or to buy them in an untraceable way for an attack.
For both the domestic and foreign market, a number of private companies are producing counter-drone technology designed to detect drones and, in some cases, knock them out of the sky. Some, like DroneShield, rely on sensors that can detect the audio signature of drones and warn police of a drone heading to an unauthorized target.181 Others, like Dedrone, are working with radio frequency (RF) sensors to detect drones in the sky, provide video and documentary evidence of the incursion, and enable counter-measures. Others are designed to knock drones out of the sky through any means necessary. The Departments of Defense and Homeland Security have also purchased one hundred DroneDefenders, “non-kinetic” rifles that disable and knock drones out of the sky.182 A British company, Selex, developed a technology called Falcon Shield which takes control of a rogue drone and lands it safely.183 The US Army is also working with companies who propose to jam the signal of the drone or disable it with lasers.184 The most widely used drone detection technique is still jamming, but doing so still presents obstacles for interdicting drones and bringing them down safely, especially in urban environments.185 Some drone producers are even experimenting with a form of in-built deterrence. In 2015, DJI, the China-based company responsible for the most popular hobbyist drones on the market, installed a firmware update to its Phantom 2 and 3 models that prevents users from flying drones near sensitive sites such as airports, nuclear facilities, military installations, and across national borders.186 This process—called geo-fencing—can be circumvented or hacked by those with advanced programming skills, but the average user will find it more difficult to fly drones in places forbidden by the authorities.
The race to find effective counter-drone technology is an important one, but it will not be easily won. An analysis by the Center for the Study of the Drone found 230 different counter-drone products available across a wide range of manufacturers and countries. But none of these are foolproof.187 Not every drone works off the same signal or can be detected by a single counter-drone system; as drones proliferate and new varieties from more manufacturers emerge, more drones will be able to evade detection and get closer to their targets. Most electro-optical (EO) and RF counter-drone systems must have a direct line of sight to the drone to jam its signal. Even so, there is always a risk of false positives if counter-drone systems mistake something benign, such as a police drone, for a threat and knock it from the sky. Even if a drone can be detected, interdiction poses a particular problem because it is hard to bring a drone down without risking harm to people beneath it. Even if that were not the case, many police forces and other law enforcement organizations lack the legal authority to bring down threatening drones in US airspace.188
This race between terrorists acquiring and deploying drones and governments finding sufficient defenses will be crucial, and perhaps dispositive, for the number and frequency of terrorist drone attacks in the future. If counter-drone technology develops rapidly into a highly effective shield against drone attacks, terrorists will gradually become deterred from seeking drones for attack and may turn to other means. If this counter-drone technology lags, or is riddled with holes, the incentive for a terrorist organization to use drones for an attack will increase. There is a risk of a cascade effect: if one successful terrorist drone attack occurs and illustrates the weaknesses in anti-drone technology, others may be tempted to follow suit and strike before the technology can adapt. These dynamics mean that governments will remain under pressure to prevent attacks and, if possible, to anticipate when these attacks are likely. To do this, they will need to turn to surveillance technology of a scope and level of sophistication never before available. And they will likely turn to drones to make sure that their societies are properly surveilled and safe, even if the use of those drones poses a risk to their democratic character.