Amid the death and destruction of the current conflict between Armenia and Azerbaijan, one shining success story has emerged: the Turkish-designed and manufactured Bayraktar TB2 drone, widely credited with stellar results against Armenian forces. ‘Thanks to advanced Turkish drones owned by the Azerbaijan military, our casualties on the front shrank,’ Azerbaijan’s president, Ilham Aliyev, boasted to the Turkish news channel TRT Haber. ‘These drones show Turkey’s strength. They also empower us.’ Few have been prepared to dispute this assertion. ‘Drone Wars: In Nagorno-Karabakh, the Future of Warfare Is Now,’ an article on the Radio Free Europe website proclaimed, quoting a military analyst’s unequivocal conclusion that the Turkish drones ‘are having a significant effect on the battlefield … We’re seeing battlefield gains for Azerbaijan that we haven’t seen in 20, 25 years now.’ A story in the Los Angeles Times from the beleaguered Armenian enclave headlined ‘A New Weapon Complicates an Old War in Nagorno-Karabakh’ featured interviews with civilians terrified by the buzz of drone engines presaging a hail of high-explosive bombs and rockets. Six months ago, Turkey’s deployment of its homegrown drones to ward off an Assad government offensive in Syria’s Idlib province had glowing reviews from commentators such as Charles Lister of the Middle East Institute in Washington DC, who wrote that the weapons had ‘transformed the strategic dynamic 180 degrees’ in Idlib in just five days. The enthusiasm is easy to fathom. Drones create their own marketing commercials in the form of the streaming video they transmit to their controllers, while images of tanks and other targets exploding can be whisked onto Twitter and other social media platforms. Armed with Bayraktar TB2s, the Azeri military has been notably energetic in this respect.
Both sides in this latest round of the South Caucasus war have attacked towns and villages, clearly targeting non-military objectives. Hundreds have been killed, and thousands more have fled: in one incident, the Azeris showered cluster bombs on homes in an Armenian-populated city. But the military efficacy of drones on the actual battlefield is less obvious. The Azeris did grind out territorial gains, but in fighting that largely devolved into First World War-type trench warfare, dominated by artillery and other traditional weapons. The fighting in Syria in the spring ended with Turkey’s drone-supported allies having lost significant amounts of territory.
In any case, the Turkish/Azeri drones’ potency would be substantially diminished if only the other side would use proven countermeasures. The accuracy of drones’ laser-guided munitions depends on an unobstructed view of the target: even smoke, easily generated, provides an effective camouflage. What’s more, remote-controlled weapons rely on an uninterrupted signal to their controller, which is eminently jammable – a tactic at which the Iranian military seems adept, having used it to capture an American ‘stealth’ drone in 2011.
The record of this year’s wars shows that although these weapons may not provide a decisive edge in combat they excel in self-advertisement, projecting an image of all-seeing omnipotence. Drones induce terror in civilian populations and healthy profits for manufacturers. The spell persists even when they are unarmed. Protesters in Minneapolis on the morning of 29 May, three days after local police murdered George Floyd, were monitored by a Reaper drone deployed by US Customs and Border Protection, circling four miles above the city. News of the deployment elicited widespread alarm; a letter signed by several senior members of Congress condemned the use of ‘live video feeds’ for the purposes of law-enforcement. All demonstrations during the post-Floyd protests were exposed to the full machinery of Trump government surveillance – intrusive monitoring of social media, cellphone traffic and cameras at ground level – but it was the use of drones that caused the outrage, a testament to the fascination these machines exert.
In fact, the Minneapolis video can’t have been of much help to government agencies, however malign their intentions. The laws of physics impose inherent restrictions on picture quality from distant drones that no amount of money can overcome. Unless pictured from low altitude and in clear weather, individuals appear as dots, cars as blurry blobs. A ‘live video feed’ sounds ominous, but the quality of the imagery depends on the quality of the cameras and the amount of information conveyed in the signal stream transmitted back to the controller. The resolution (sharpness of detail) of the image sent by the drone – in any case inherently far worse than video – is dictated by the size of the radar antenna, which is limited by the small size of the aircraft. In 2010, a US air force drone crew watching infrared video of a night-time convoy of Afghan vehicles, as revealed in the transcript of their conversation, thought that warm dots in the trucks they saw on their screens were weapons. In fact, they were turkeys, presents carried by the peasant passengers for their relatives in Kabul. (The trucks were attacked, killing 23 people.) Even in daylight, distinguishing children from other potential targets poses particular difficulties. An air force pilot once told me he had spent an afternoon circling high above southern Afghanistan, watching four tiny stick figures on his cockpit video screen. They clustered for some time at the side of a road in a manner that suggested a Taliban bomb-laying party, which would qualify them for instant incineration with a Hellfire missile. But as he continued to watch they retreated across the fields to an isolated farmhouse – at which point a visibly taller figure emerged to hustle the children inside for supper.
Such technological realities seldom intrude on popular faith in drones, and are hardly challenged in the books that have been written about them. Michael Boyle describes Gorgon Stare, an array of video cameras designed for the Predator B drone, a large-wingspan aircraft that operates at an altitude of 25,000 feet, as offering ‘persistent, wide-area surveillance of small towns’, enabling intelligence analysts to track the movements of malefactors. But the air force unit assigned to test the system in 2012 was less impressed. Alongside its derogatory conclusions, their report included a pair of high-altitude photographs of the unit’s own base. One had been taken by Gorgon Stare, developed at a cost of $500 million to the taxpayer. The other, identical in quality of detail, had been downloaded from Google Earth, free of charge. In neither were humans distinguishable from bushes. Thomas Stubblefield writes excitedly of Argus, an even more technologically ambitious system that permits users to ‘zoom into pedestrian traffic on a given street, follow a vehicle of interest, or even map the audio of intercepted telephone calls spatially onto this representation of the city’. He ruminates on the profound implications of a technology that can, by cross-referencing historical patterns, provide ‘limitless temporal parameters’. But when tried in Iraq, reviewing past drone video of sites of insurgent bomb attacks in the hope that it would reveal perpetrators in the act of planting bombs turned out to be fruitless.
It may seem unnecessary to belabour such shortcomings, especially in view of the ease with which commercial drones – $80 for a simple quadcopter from Best Buy, or £50 from Argos in the UK – have put high-quality aerial views of the landscape in reach of any amateur filmmaker. Low-altitude drone video has been a boon in carrying out many important activities, from fire-fighting to pipeline maintenance. And small, cheap, off-the-shelf drones have evened the score somewhat for guerrilla operations, as demonstrated by Islamic State in their deployment of quadcopters adapted to drop grenades on US forces in Syria. The Yemeni Houthis went a step further, mass-producing their own drones using a 3D printer supplied by the Iranians.
But the mystique of these unmanned machines ultimately derives from their presumed ability to gather information, enabling the selection and destruction of discrete targets with unique precision at the click of a button on the other side of the world. In aspiration as well as consequence, there is nothing fundamentally new here, since similar claims have been advanced on behalf of strategic bombing for a hundred years, with invariably disappointing results. Bombers dispatched by the infant RAF in 1919 to eliminate the ‘mad mullah’ in British Somaliland took their orders directly from London. But their best efforts, a local British army commander subsequently recorded, left the mullah’s supporters ‘cheerful, utterly defiant, and grossly slanderous about my parentage’. The US strategic bombing campaign in the Second World War was premised on the theory that targets vital to the enemy’s war effort could be destroyed by bombers flown from distant bases, striking their targets with unerring precision thanks to new technology. But results in combat were deeply disappointing: the bombs almost always missed by a wide margin. And the belief that the enemy’s war economy was analogous to a mechanical device that could be disabled by the elimination of crucial components, such as Germany’s ball-bearing factories, turned out to be wrong. Since the staggering cost of the bombing effort had to be justified and the myth of bombing as a war-winning tactic preserved, the unachievable goal of precision strikes was ultimately abandoned in favour of indiscriminate attacks, most dramatically the incineration of 300,000 civilians in the fire raid on Tokyo in March 1945.
Nevertheless, the dream of remote precision endured, along with the belief that an opponent’s operational system could be comprehensively understood and selectively dismantled. During the Vietnam War, civilian scientists in service to the Pentagon devised an ‘electronic fence’, made up of thousands of sensors dropped across the jungles of North Vietnam. These were designed to detect enemy troop movements by such tell-tale signs as the smell of urine or ground vibrations from the movement of trucks and tanks. But the North Vietnamese soon devised effective countermeasures, depositing buckets of urine far from their troops in locations that were then duly bombed while supplies continued to flow undetected along the jungle routes from north to south. The super-secret operation was ultimately revealed in the Pentagon Papers, leaked by Daniel Ellsberg. Even though it had proved to be a fiasco, the notion of a remote-control killing machine caught the public imagination, symbolising the soulless nature of the American war effort. One embittered veteran, Eric Herter, spoke eloquently at a Boston anti-war meeting in 1971 in terms that would be echoed by veterans and other critics of the drone wars fifty years later: ‘This new war … will be a war not of men at arms, but of computers and weapons systems against whole populations. Under its auspices, the people of the villages have gone from being “gooks” and “dinks” to being grid co-ordinates, blips on scan screens, dots of light on infrared film.’
The failure of technology to deliver results did not lead to a fundamental rethink of American strategy. Instead, the preferred belief was that improvements in technology would in the end do the trick. This article of faith seemed to be vindicated by videos of bombs and missiles unerringly striking their targets during the 1991 Gulf War, though an exhaustive investigation by the Government Accountability Office revealed that the attacks had been markedly less effective than military publicists had suggested.
Advances in sensor technology, however inadequate, may have improved the pictures enough to induce the trauma experienced by some drone crews in recent years. In Hellfire from Paradise Ranch, Joseba Zulaika relays the harrowing reminiscences of drone operators, their victims obliterated in an instant by Hellfire missiles fired on command from a trailer in the desert outside Las Vegas. Like others before him, he is intrigued by the contrast between drone crews’ daily occupation – staring at endless images of landscapes and people in the Third World, sometimes killing the latter – and their mundane domestic existence once they step out of the trailer and back into middle-class America, perhaps stopping on their commute home to pick up milk for their children from the grocery store. But the anguish expressed by the former crewmen he quotes may not be widespread in their community. When the commander of the team that orchestrated the killing of those 23 turkey-bearing Afghan civilians denied that his men had been ‘out to employ weapons no matter what’, an investigator responded that they had expressed exactly that intention at least 14 times in the course of the mission. What’s more, the sensor operator had complained about an inconvenient report that there might be a child aboard one of the targeted vehicles, thereby possibly impeding a kill shot. In almost twenty years of drone assassinations there is no confirmed record of any drone crewmember ever refusing an order to kill, though some half a dozen, including those quoted by Zulaika, have spoken publicly of their subsequent remorse and symptoms of PTSD. The US air force has trouble retaining drone operators, and has promised more psychological counselling, but most complaints appear to be about the irregular hours that result from the constant shift changes ordered by headquarters.
However many drone operators actually suffer from ‘moral injury’, defined in an article in the Journal of Clinical Psychology in 2009 as ‘perpetrating, failing to prevent or bearing witness to acts that transgress deeply held moral beliefs’, the affliction evidently doesn’t extend far up the chain of command, where the ability to order politically risk-free assassinations holds great attraction. ‘Turns out I’m really good at killing people,’ Barack Obama remarked on the day when Anwar al-Awlaki, an American citizen, was executed by drone at his command. ‘Didn’t know that was going to be a strong suit of mine.’ Zulaika reminds us just how avidly Obama took to the business, even boasting in a radio interview of how he ‘took out’ a Taliban leader, without mentioning that the man’s wife had been with him at the time.
This was one element of his predecessor’s legacy that Trump was happy to preserve and enhance, killing more people in drone strikes in the first year of his presidency alone – including 250 children in Pakistan and Yemen – than Obama had managed in eight. Obama milked the killing of Osama bin Laden for political advantage, and Trump has similarly revelled in drone assassination as an extension of publicity by other means. He reportedly insisted on killing Hamza bin Laden, a son of al-Qaida’s founder but relatively unimportant in the organisation’s hierarchy, solely because of his famous name. Trump also commissioned and celebrated the killing of Qasem Soleimani, the leader of Iran’s Quds Force, and eliminated the IS leader Abu Bakr al-Baghdadi in a land attack. As Boyle correctly points out, the real danger of the drone programme is that it carries few political risks and so tempts leaders into casual acts of war.
These tendencies have been vividly displayed by Turkey’s president, Recep Tayyip Erdoğan, whose drones – designed by his son-in-law, Selçuk Bayraktar – have carried out innumerable assassinations, not only among the leadership of the insurgent PKK and the civilians around them, but also of senior foreign officials: Turkish drones killed two Iraqi generals last April. Erdoğan clearly believes that his weapons, one of which flies with his personal presidential signature on its bodywork, ‘show Turkey’s strength’, in the words of his Azerbaijani ally Aliyev, without running the political risk of too many dead Turkish soldiers. Turkey is only the most prominent example of drone proliferation, as an increasing number of countries invest in their own fleets of aerial robots, a trend that will surely encourage irresponsible military adventures, not to mention assassination as an instrument of state policy.
The flaw in this approach is that robots don’t necessarily do the job. If they do, the results tend to be unpredictable, and very often unpleasant. The strategy of ‘high-value targeting’ has shifted focus from things – those ball-bearing plants – to people, but the results have proved no more satisfactory. US drug enforcement strategy has long been focused on capturing or killing the leaders of drug cartels, the ineffective consequences of which are unhappily reflected in an ever increasing flow of narcotics to American consumers. The elimination of insurgent jihadist leaders in Iraq in the years following 2003 immediately boosted attacks on American forces as new and more aggressive leaders took command. The pattern was repeated in Afghanistan, as the killing of Osama bin Laden was followed soon after by the emergence of IS. The assassination of Soleimani may have been welcomed in Tehran, given his record of alienating Iran’s supporters in neighbouring countries, and the death of al-Baghdadi had little effect on IS’s guerrilla campaign.
But there is little prospect that the strategy will change any time soon, for reasons that are well described by Christian Brose, though perhaps not in the way he intends. Brose, unlike the academic authors of other recent books on drones, is a denizen of the defence complex. Formerly staff director of the US Senate Armed Services Committee, which oversees the military budget, he is currently employed by Anduril Industries, a defence contractor spawned in Silicon Valley. His book reveals him as a type of figure common enough in American defence-intellectual culture: the self-proclaimed ‘maverick’ who in reality is a staunch defender of the status quo. Starting from the proposition that the US military is falling seriously behind in military competition with China, he derides ‘legacy platforms’ such as the $13 billion aircraft carriers cherished by the US navy, and advocates shifting to a force attuned to revolutionary technologies.
Drones feature heavily in the scenario Brose lays out in The Kill Chain, including prospective designs that could be controlled by ‘neural signals’ relayed directly from the brain, ‘not just one, but groups of them’, making it possible for human beings to ‘direct and oversee the operations of drones and other robotic military systems purely with their thoughts’. Beyond that, he invokes the alluring prospect of ‘autonomous’ drones that operate independently of human guidance thanks to the wonders of artificial intelligence. New and exciting though this – along with other less fanciful but equally unlikely concepts promoted by Brose – may sound, it is entirely in keeping with the rules of classical budget enhancement as practised in the US defence complex. These require in the first instance the egregious exaggeration of a prospective opponent’s capabilities. Thus China – despite currently being incapable of manufacturing its own advanced integrated circuits, foreign supplies of which have been blockaded by Trump – is deemed to be racing ahead in the development of advanced military technology of all kinds, including drones. Meanwhile the US, held back by entrenched bureaucratic interests, is not realising the potential of the technologies available to it, a disaster in the making that can be averted only by putting money in the right hands.
Brose’s novel proposal for overcoming the resistance to his reforms by entrenched interests is to enlist the support of defence industry lobbyists – the praetorian guard of those interests – with promises that their clients will make just as much money with the new programmes. Throughout, he pays fulsome tribute to his late employer, Senator John McCain, a prominent exemplar of the ‘maverick’ syndrome, prolix in eloquent denunciations of sacred military cows, while never taking any practical action to discommode the Pentagon on any important matter – and certainly not in the invocation of China as an imminent and terrifying threat. Occupying a scarcely less prominent position in Brose’s pantheon is the late Andrew Marshall, who was for decades the director of the Pentagon’s Office of Net Assessment. Marshall was an immensely powerful bureaucrat who enjoyed a reputation for unorthodox thinking, yet nevertheless somehow always wound up making the case for increased defence spending. Following the supposedly decisive triumph of precision weapons in the Gulf War at the turn of the 1990s, Marshall and his adherents in the defence think tanks he was so lavishly funding assiduously promoted the notion of an ongoing ‘revolution in military affairs’ that would at last realise the dream of those early bombing campaign planners. It was the funding doled out by Marshall that led to the production of drone weapons that could pick off individuals.
The overall result of these developments has been to create an ever widening distance between those in charge and the real world. Live video feeds on the desks of four-star generals may give the illusion that they are conveying reality, but they are two-dimensional images synthesised by a preprogrammed centrally controlled system ever vulnerable to opponents’ unanticipated actions (those diversionary buckets of urine under the jungle canopy). But a change of course is vanishingly unlikely. For the presiding authorities, in and out of uniform, the principal if unspoken attraction of the current system is its enormous cost and consequent profitability, not to mention the sense of empowerment it affords. To take too clear-eyed a view of the world would be dangerous.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.