I began thinking and writing about drones around 10 years ago, during my master’s degree. At the time drones seemed like the single most dramatic warfighting breakthrough we’d seen in decades. At this time, though, there were only really a couple of kinds of drones being used on battlefields. General Atomics’ Predator and Reaper drones were the missile delivery systems, while Northrop Grumman’s Global Hawk – boasting a flight duration of more than 30 hours – was the big game in town for always-on surveillance.

Those drones were the perfect enablers for the wars the US and its allies were fighting in the Middle East. An asymmetric conflict – where one side has vast technological superiority over the other – meant that the skills of fighter pilots weren’t as necessary to contest terrain. And the hotly contested politics of the Iraq and Afghanistan conflicts meant that any Western fatalities in combat would be harshly criticised. To avoid flag-draped coffins returning to our shores, better surely to fight with machines rather than soldiers?

Inevitably, Russia’s invasion of Ukraine has refocused the debate about European defence. Investors have jettisoned principles they held sacrosanct a few years ago, leading to surging stock prices in established national champions like Rolls-Royce and Rheinmetall and more cash for ‘disruptive’ startups like Helsing, which raised $450 million in venture funding last year.

There is a complex and relatively open customer base for startups building in the defence supply chain, whether it’s parts or software. But startups like Helsing, that produce weapons, find themselves sandwiched between two extremely powerful forces. On one hand, the incumbent defence champions enjoy some of the most tightly-knit government relationships anywhere. And on the other, there is new bottom-up pressure created by the surprising effectiveness of cheap improvised drones, seen in action over the last three years in Ukraine.

What is Europe good at? Building very high-quality products, with the best materials and with the highest standards of manufacturing integrity. But quality, in a kinetic conflict, matters much, much less than efficacy.

The drones in use on the Ukrainian frontline vary from genuinely best-in-class to improvised devices that are essentially rejigged hobbyist drones. This duct-taped drone army is in fact far more customisable and repairable than expensive kit from manufacturers like Helsing, and can in many cases be re-programmed and repurposed by combatants in the field. Not to mention, they are far cheaper than even the most cost-effective European alternatives. While some drones on the Ukrainian frontline can cost as little as €360, Helsing’s lowest-cost HF-1 drone is priced at around €16,700.

Other governments are advocating their own solutions to the new unmanned warfighting paradigm. One newer high-performance entrant in the drone market is the Bayraktar TB2, made by Turkish national champion Baykar, which can cost less than $5 million per drone (compared to around $20 million for a US-made Predator). The TB2 is undercutting US and European high-end products while being explicitly backed by the Turkish government, which is itself becoming the world’s fastest-growing national military-industrial complex.

And so Helsing appears to be caught in a massive, unpredictable and rapid technological deflation. Frontline conflict doesn’t actually need expensive surveillance drones – after all, your enemy’s bullets are flying right over your head. The biggest competition to the weapons being manufactured by Helsing (and American challengers like Anduril) right now, at least on the Ukrainian frontline? A £1,000 DJI filmmaker’s drone with a bomb stuck to it. Meanwhile, the elite drone market is being locked up by national champions and defence primes with decades of experience winning government contracts.

Weapons production is only one part of the defence sector, of course, and Helsing has already shown that it is able to secure valuable contracts such as real-time protective analytics for the ongoing Eurofighter upgrade. But the companies that choose to build weapons have opted into another level of moral and journalistic scrutiny – which is exactly as it should be, given the products these companies are building will kill people.

Time will tell whether Helsing’s progress will continue. But right now, its attempt to build weapons of war with software-style margins makes for a shaky investment thesis.

But at the same time, I’ve let 18 months of development in unmanned weapons slip by without comment. I find it nigh-on impossible to confront the situation in Gaza but there are interesting recent developments in the unmanned air war in which Russia and Ukraine are currently engaged.

Some of the most consistent laws of technological development are playing out in the drone market. As smaller chips start to pack more punch for manufacturers, they’ll inevitably seek out lower-cost alternatives to their existing product suite. Indeed, sometimes it’s felt as if over the last few years, the race in drones has been to push the smallest systems possible on to the battlefield, to offset the punishing cost of the most powerful weapons systems. (As covered in my last post on Ukraine, using a single Javelin missile costs $78,000.)

Using the biggest, most expensive munitions – manned or unmanned – is harder to justify when it’s two neighbouring countries fighting each other. While the US ‘needed’ the $220 million Global Hawk drone to manage its conflicts in the Middle East (and to guard against anything kicking off in the South China Sea), Ukraine does not need the same level of sophistication to monitor the Kursk region. That’s part of the reason why manufacturers like Aerovironment have seen higher international demand for its much daintier unmanned systems in the last year or two.

Aerovironment is the maker of the well-established, and teeny tiny, Switchblade 300 ‘kamikaze drone’, which debuted in 2011 and is still a big seller. The latest version of the 300, dubbed Block 20, weighs somewhere around 2.25 kgs, which is truly miniscule. But even specialists in smaller weapons are identifying a need for greater firepower. Aerovironment’s much newer Switchblade 600 (originally released in 2020) has now made its way to the battlefield in Ukraine, if reports are to be believed.

The Switchblade 600 has different use cases from the 300, of course: it’s an anti-tank warhead with capabilities that the 300 could never hope to emulate. Around 10 times heavier than the 300, the 600 can stay in the air for 40 minutes or more, compared to the 20-minute flight time of the 300’s Block 20 iteration. (I’ll let you reflect on whether a munition system that has a 20-minute flight time can reasonably be described as ‘loitering’. Could a teenager doing wheelies outside a newsagent for 20 minutes be nabbed under the same charge?)

There is room for a separate post on how we define loitering munitions vs ‘drones’. One simple (and therefore reductive) definition is that a drone fires missiles, while a loitering/suicide/’kamikaze’ munition is the missile. It identifies, tracks and destroys a target itself. The militaries using them laud them as being more precise, because they cause less widespread collateral damage due to their size.

Definitions aside, the earliest use I can find of the Switchblade 600 in Ukraine was in October, when essanews.com, a division of the broadly respected Polish media network WP, reposted footage from social media indicating that a Switchblade 600 had destroyed a Russian air defence system in the Donetsk region. There are suggestions from dubious outlets that the Switchblade 600 may have its own vulnerabilities*, but the recent success indicates that the Switchblade 600 is “highly effective” against even the latest Russian ground systems.

That potential success may have helped validate the US government’s decision to include the Switchblade 600 in its fairly new Replicator programme – an initiative designed to speed up the commercial manufacture and deployment of mission-critical autonomous technologies. In June this year, it was reported that the US Army has committed to buy more than 1,000 Switchblade 600s.

While the tiniest drones still do useful stuff for fighting forces, it takes a certain amount of firepower to make a difference when you’re up against sophisticated anti-drone systems. Here’s hoping it won’t take 18 months to get another blog post into your inboxes.

—

*Just this week, EADaily, an outlet for Russian propaganda, reported that a Switchblade 600 was shot down “with a well-aimed shot from a gun”. If there’s any truth in that, well, chapeau.

The live CIA demo astutely sums up one of the key truths of complex AI systems: that when the context shifts and edge cases appear, even the most sophisticated algorithm is liable to misread their environment. In this case, a piece of debris jettisoned after Newton blows up an empty building lands in a heap that coincidentally gives it a quasi-human face. Newton frantically scans its database, trying to match the debris with one of the thousands of deceased people in the CIA’s files, to no avail:

The apparent civilian death he’s caused sends Newton into a metaphorical – and later literal – tailspin. Despairing, Newton abandons his official mission and attempts to seek solace by connecting with people. After he lands on Main Street in a remote town, the local people daub Newton’s fuselage with a smiley face.

As Newton continues to circle and as his fuel dwindles, he decides that he would be better off destroying himself, fitting given his purpose as a machine designed to destroy. His ‘suicide’, and a final pitiable smile, provides the film’s denouement:

DRONE asks us to consider whether machines can ever have ‘souls’, or, at least, intentions of their own. It is a timely question when just a few months ago, the CIA had to deny conducting simulations where drones killed their operators after the US Air Force’s Chief of AI Test and Operations alleged at a conference that a test did indeed go awry in this way.

The USAF staffer in question, Colonel Tucker ‘Cinco’ Hamilton, appears to be highly concerned about machines operating autonomously and taking decisions for themselves on the battlefield, enabled by AI. Any weapons system able to decide whether to pull a trigger is inherently ethically dangerous. But war comes in many forms, and in some theatres, we have more clarity than others.

The prospect of nuclear war is taken more seriously today than it has been in decades. It appears that the US and UK have resolved to keep humans in the loop throughout nuclear decision-making processes, as and when AI starts to become integrated with these command-and-control systems.

The dangers of nuclear war have been burned into our brains over decades. The dangers of drone warfare, in contrast, are still kept from the average Western consumer of news. The reason seems to me crystal clear: nuclear armageddon necessarily touches all our lives, while up until now, drones have been something that happens to other people, mostly bad guys (if you believe the military spokespeople).

But drones are becoming part and parcel of modern warfare. They are a critical asset for both Ukraine’s and Russia’s militaries. They were pivotal in shaping the Nagorno-Karabakh conflict, and the radically different unit economics of drone warfare are turning countries like Turkey from relative also-rans into real military power brokers.

Buckelew’s film reflects the new primacy of drones in combat. Using drones means no loss of human life on your side. Drones are also fantastically cheap compared to the cost of a fighter jet or manned submarine. And because it’s so difficult to determine culpability when things go wrong, drones are extremely useful for politicians and military leaders seeking to evade accountability wherever possible.

So it’s no surprise that in DRONE, the CIA are excited about Newton’s potential to fight wars more intelligently, more proportionately, even more humanely, than ever before. This is exactly the message on drones that governments want us to take away: that they’re just a smarter, cleaner way to fight.

But as we race to insert AI into all manner of military processes and control systems, DRONE serves as a reminder that autonomous machine may make unexpected decisions. In suffering a crisis of conscience that amounts to an awakening, Newton reminds us that in a military context, moral clarity is a vice, not a virtue. The more deep, complex technologies we embed into military decision-making, the easier it becomes for militaries to obfuscate and evade true accountability.

Technologists are used to winning big markets. But the only one of the above markets where tech has really made a dent is in retail. A big chunk of that juicy $800 billion defence budget is reserved for technological ‘innovation’, but most of the big contracts are still carved up between the traditional defence primes like Lockheed Martin, Raytheon and Boeing.

Sounds like a market ripe for disruption! Which brings us to Anduril. Palmer Luckey was very happy building VR headsets at Oculus, until he was booted out a few years after Mark Zuckerberg spent $2 billion buying him out and stuck him in a basement at Facebook. So he decided that the next step would be to turn his engineering prowess to building military intelligence systems.

He’s done this quite successfully: bucking the tech financing landscape, Anduril announced a new $1.5 billion funding round at a valuation of $8.5 billion this month. But Anduril exhibits many of the usual contradictions, elisions and hypocrises that characterise corporate communications in the defence industry.

Most defence companies say that their goal is to make conflict safer. Which is a bit like Shell saying its goal is to move past fossil fuels: meaningless. CEOs and founders lean on a language of deterrence to assuage fears. Luckey borrows this template in a Wired interview, explaining the benevolent reason why Anduril exists:

That’s the thesis of our company: You want to have really strong technology that deters conflict by raising the cost high enough so that it's not thinkable.

“Raising the cost high enough” essentially sets Anduril’s stall out as a future creator of truly world-destructive weapons systems. It’s also an interesting phrase because elsewhere, Anduril rips off the standard ‘tech disruptor’ template by saying how cheap their new technology will make it to wage war. As Luckey’s co-founder Brian Schimpf says in a 2021 Defense News opinion piece:

With a radically reduced manpower constraint, the United States could discard its exquisite systems in favor of swarms of cheaper, artificially intelligent alternatives, which could conduct a range of missions with far less human oversight.

…cool? In the same piece Schimpf further elaborates on the parlous state of modern warfare:

Our current unmanned systems are, in truth, unmanned in name only, often requiring large crews of remote pilots and operators.

Not really sure what the problem is here tbh Brian. If we are going to take seriously an argument that shortening the OODA (Observation, Orientation, Decision, Action) loop is always good, it’s probably worth reiterating two reasons why having a ‘human in the loop’ is useful.

• Firstly, the principle of accountability is central when something goes wrong with one of these systems. You cannot punish a machine for committing war crimes. And manufacturers are rarely, if ever, held liable. So with an autonomous machine, who gets the blame when something goes wrong? The blog Rethinking SLIC has a very useful summary of the liability gaps that emerge when remote, AI-enabled or autonomous weapons systems are in use.

• Secondly, even the most powerful AI systems have been unable to replicate the nuanced decision-making of a human assessing a conflict zone. Some analysts use straw man examples of fictional dogfights to laud the advantages of shortening the OODA loop, but there has been no evidence that autonomous systems are anywhere near being able to discriminate between combatants and non-combatants; failure to adequately do so in conflict is a contravention of the Geneva Convention relating to the protection of civilians in war.

Anduril isn’t the only company seeking to apply Silicon Valley principles to a huge global market. Helsing, backed by Spotify founder Daniel Ek, has just partnered with Rheinmetall to “provide the armed forces with advanced and future-proof capabilities”.

Luckey views the development of autonomous weapons systems as inevitable in the face of the West’s big arms race with China and Russia. His argument is essentially, ‘If you don’t build it, they will, and we’ll lose.’ In one sense, he’s right: China and Russia will definitely try to build more autonomous systems. They’re abetted in this by a confluence of two megatrends which, combined, pose a perilous threat to the world order:

• Tech is the only show in town for a global economy which looks weaker than the fundamentals. We have experienced a huge global bull run lasting more than a decade. Desperate to avoid a slowdown, the biggest tech companies will push for further deregulation and defy their egalitarian founding missions to push into defence as one of the only trillion-dollar markets left to be disrupted.

• Since the formation of the United Nations post-WWII, respect for, and trust in, multinational institutions has never been lower. There is simply no chance that China, Russia and the US will slow down when there are so few roadblocks to autonomous weapons development and nuclear rearmament. The territorial integrity of smaller nations hasn’t been this vulnerable in 30 years.

I’m very worried about a vacuum of accountability in defence, which leaves the door open for incumbents and upstarts alike to build new, destructive autonomous systems that actually lower the bar to conflict. Companies like Anduril are accelerating us towards a hair-trigger environment where the actions of robots could tip us into kinetic conflict. This is far from the deterrent Palmer Luckey seems to think it is.

More than ever, the sky is a contested space. Beyond the early commercial and military applications of air travel, recent decades have seen the sky start to host near-infinite webs of communication and surveillance technologies. Today, whoever controls the sky controls a uniquely effective stratum for intelligence gathering.

Illingworth is worried about what this means for the earth-bound billions who are anxiously looking upward or altogether ignorant of what’s happening above our heads. Topologies of Air stitches together informal collages of material and imagery with longer, richer film and video installations. The thread running through each work is the chemical, geopolitical and military transformation of the sky into an informational and strategic asset.*

When we think about the sky, we think about ideas of openness and freedom. But the sky is no longer a free space. In particular, the colonisation of the sky turns the activities and behavioural patterns of those on the ground into a new resource for those who control the air.

Arguably, the exhibition’s centrepiece is ‘Airspace Tribunal’, a series of hearings convened by Illingworth and barrister Nick Grief, who works alongside Illingworth at the University of Kent. The stated goal of the Airspace Tribunal, which has its own website, is securing ‘A New Human Right to Protect the Freedom to Live Without Physical or Psychological Threat from Above’.

The hearings themselves are gruelling examinations of people’s experiences at the mercy of airborne warfare. Citizens from different conflict zones lament the missiles that one day speared out of the sky, mutilating their friends and destroying their communities.

The hearings form a powerful commentary on the essential asymmetry created by the control and deployment of the sky as a strategic asset. If an object is high enough in the sky, it is invisible to those on the ground. It is extremely easy for this effect to generate a state of near-terror for the people surveilled by aerial hardware or software. In her testimony to the Airspace Tribunal (transcript available here), neuropsychologist Catherine Loveday attests that “an unseen threat is always going to have a more powerful effect.” Constant airborne attacks inflict psychic damage on populations, with no discrimination between civilians and combatants.

In ‘Airspace Tribunal’ Illingworth places footage of the hearings themselves alongside dreamy skyscapes, sharpening the contrast between the gruesome realities of aerial warfare and the artist’s utopian vision of a sky free from interference.

Is an examination of modern airpower in conflict enough to deliver real change, though? I was left wondering whether citizen-led movements and academic debate will ever be enough to overcome the massed ranks of aerospace and military strategists (not to mention the politicians hoping for juicy non-executive directorships on their retirements) who see the sky as a resource too valuable to surrender.

But ‘Airspace Tribunal’, and Topologies of Air in general, is worthwhile because of the white-knuckle testimonies that bring home the peril of living under a sky that can at any point morph into a battlespace, with missiles suddenly screaming out of cool blue. The exhibition feels like an elbow in the back for Toulouse, a beautiful, relaxed European city that happily hosts many builders of the hardware and software that power our modern conflicts. It’s not going to turn the tide on its own, but Topologies of Air is a helpful provocation.

—

* The sky is not the only venue for this kind of technological and philosophical conflict. Illingworth’s work brought to mind Martin Arbeloa’s Planetary Mine, which investigates “the synergistic effect of innovations in robotics, biotechnology, artificial intelligence, and geospatial information [that has] exerted a fundamental overhaul in the extraction and processing of minerals.” Arbeloa is concerned about what lies beneath the earth’s surface, rather than the skies above. But Planetary Mine and Topologies of Air share aspects of the same story: disruptive technologies enabling groups to ‘claim’ rights over areas of the earth that were previously outside any human dominion. You can get 40% off your copy of Planetary Mine with Verso now!

Some shipments of Javelin missiles have arrived in Ukraine without phone numbers for US technical support, and there are reports that the same shipments could also have been missing the computer training programs required to get soldiers ready to use them in combat.

A Javelin launcher system costs around $178,000, and each missile launched costs an additional $78,000.* The US is already depleting its own supplies of Javelin and Stinger missiles – around a third of all stocks have been sent to Ukraine since Russia invaded. Any ‘wasted’ missile shipments will increase the pressure on the US to rationalise its spending, which is already starting to worry some lawmakers. So in more ways than one, gaffes like this aren’t ideal for Ukraine, which is already worried about the conflict becoming “normalised” in the minds of allies watching on around the world.

In any event, when it comes to drones in Ukraine, it seems to be a case of the cheaper, the better. Riller Kobots has already written about how Turkish Bayraktar drones are proving to be the stars of the show in cheaply and effectively arming Ukraine against Russia’s offensive. Other cheap weapons systems, such as Switchblade 300 drones, are in their own way showing themselves to be just as useful as Javelins.

Dubbed ‘Angry Birds’ in some quarters, Switchblades are 2.5 kilogram kamikaze weapons that (when directed) launch themselves at targets to cause damage.** Evidence suggests that the 300s are effective but only when aiming for smaller and less significant targets. That’s why the US is building the Switchblade 600, which is around 10 times heavier and is being billed as a ‘tank buster’.

All well and good, as long as the US actually provides numbers to its call centres when these bigger and more complex weapons systems aren’t working properly.

Meanwhile, everyone knows that the most cost-effective way to get a working drone capable of causing damage is to sellotape a bomb to a £1,000 DJI drone. (This is what lots of Ukrainian IT professionals are spending their time on.) For now, the ingenuity of Ukraine’s volunteer combatants is making up for faulty global supply chains and absent-minded logistics professionals. How long this jury-rigged fighting force can hold out in the face of Russian territorial gains remains to be seen.

—

*The Drive’s analysis of the Washington Post’s original reporting quotes a higher figure for the cost of a Javelin missile, but its original source appears to corroborate the $178,000 figure I cite above.

**Switchblades have proven so successful in Ukraine that France now wants in on the act. Colonel Arnaud Goujon wants to show he’s no chicken by deep-frying France’s enemies with some new munitions.

Since 9/11 conflict has been defined by global powers, such as the US and UK, battling insurgent forces of non-state terrorist organisations and ragtag militias. In Ukraine, a global power is the insurgent force, and in the conflict’s first 50 days Russia has been frustrated by defiant Ukrainian forces using everything they have to retain territory and repel surprisingly disorganised Russian troops.

Without the military budget of a global power, and with Western allies unable to provide direct defence support, Ukraine can’t take its pick of the most expensive and sophisticated drone technology. But unmanned aerial vehicle (UAV) manufacturing is a much more globalised game than it was in the aftermath of 9/11. Turkish Bayraktar TB2 drones have been at the centre of Ukraine’s remarkable feat of counterinsurgency.

Now, these drones aren’t as fast, stealthy or deadly as the Predator and Reaper drones made by General Atomics. They can’t stay in the air as long either, which means the base and the operator usually have to be relatively close to the front line. But there are some big plus points to these cheaper drones too, particularly in a defensive context. They’re useful in two main ways:

• As one component of a force’s toolkit, they are essentially really annoying and thus very good cover for other more powerful munitions. It is (almost) impossible that Russia’s flagship Moskva was sunk by a Bayraktar TB2,* but this article by David Hambling indicates that TB2s played an important role in the effort to disrupt and take down the warship.

• Because they are more expendable, cheaper drones can also be used in riskier situations, meaning a greater advantage gained per successful mission.

Bayraktar drones are at the vanguard of an increasingly competitive and globalised drone production industry. Baykar, the manufacturer of Bayraktar drones, is a key part of the Turkish’s government’s massive success in defence exports, which increased 7x between 2006 and 2021. The US, Israel and China no longer have the hegemony over combat drones that they once enjoyed. Baykraktars also featured in the conflict between Armenia and Azerbaijan last year.

Is there a good comparison to make between Bayraktar’s success and another industry? Cars are pretty difficult to make and kill a lot of people, so maybe the automotive sector makes sense. Bayraktar’s drones could end up working a bit like Maruti Suzuki’s cars in India. They’re not ludicrously cheap, but they’re definitely good value and they sell like mad. Each of the top four best-selling cars in India is made by Maruti Suzuki, and around 20,000 people buy a Maruti Suzuki Swift every month.

So when a piece of kit works well and provides good value, people take notice. And as Bayraktar drones continue to play an important role in different conflicts around the world, more countries will start to take notice. Drones are no longer monopolised by global military powers. This is a relief to those watching Ukraine defend itself. But if drones become a trivial cost for militaries to bear, it potentially presents a lower bar to initiating conflict. This is a worrying trend that’s worth keeping an eye on in a new era of geopolitical uncertainty.

—

*If a Bayraktar missile had hit one of the Moskva’s anti-ship missiles, it could have triggered an explosion big enough to sink the ship. But this would require Luke-Skywalker-hitting-the-Death-Star accuracy, and there is no evidence that this happened.

Joe Biden announced in February that a raid on a compound in Syria led to the death of Abu Ibrahim al-Hashimi al-Qurayshi, leader of the Islamic State since 2019.

The operation was remarkable in its similarity to the raid on Osama bin Laden’s compound in 2011. Rather than conducting an airstrike or unleashing a drone, the US sent in special forces troops that engaged in firefighting before al-Qurayshi detonated explosives that killed himself and members of his family.

Any military incursion onto foreign soil carries diplomatic and political risks, as well as directly endangering the personnel carrying out the mission. Surely it would have been cheaper and easier to carry out the mission remotely, discharging a weapon from thousands of miles away safe in the knowledge that no troops were in danger of losing their lives?

In fact, al-Qurayshi’s death shows exactly why remote warfare has limits. The involvement of ground forces improves intelligence gathering, relations with local populations, and leads to more lives saved.

First, think back to the bin Laden raid. Without soldiers there to record and testify to his death, how willing would we have been to believe that the orchestrator of 9/11 was really taken out? Without people on the ground, it is far harder to confirm specific kills of high-value targets. After the al-Qurayshi raid, his body was identified with fingerprints and DNA analysis – very difficult processes to carry out if all you have is a fighter jet.

Conducting raids with soldiers is also a way to establish a direct dialogue with local people. The Guardian’s live blog mentions the following report: “Residents who lived close to the house told Omar Nezhat, a local journalist, that they were interrogated by the soldiers, who marked their foreheads with numbers and told them not to worry, because American forces were there to kill an ISIS leader.”

Whether or not you agree with the US military’s approach to its activities in the Middle East, locals surely benefit from having real people providing information to them on the ground, rather than terrifying and arbitrary explosions from drones or jets.

Lastly, it is worth assessing fatalities and lives saved. President Biden said in his address after the al-Qurayshi raid that the mission was designed to avoid an airstrike that would have “destroyed the entire building”. al-Qurayshi lived at the top of a building which housed oblivious civilians on the ground floor. Any warhead fired from a fighter jet or a drone would have guaranteed many civilian casualties. As it transpired, commandos managed to evacuate a number of civilians from the building before al-Qurayshi set off the explosives that killed him and family members. Needless to say, a missile launched from a drone without notice cannot take account of these externalities, and cannot respect the lives of people physically close to the target in question.

Some civilians were injured and killed during two hours of combat prior to to the US troops breaching the compound. But from an ethical perspective, these deaths are easier to justify than if they had happened in a drone strike. It is accepted that soldiers in the midst of a battle (whether US or Islamic State) may take actions that leads to innocent deaths. Those soldiers must live with the consequences of their actions, having seen blood and pain up close. The experience of these combatants is markedly different from a drone pilot thousands of miles away from the battlefield, unable to witness the immediate and visceral consequences of their actions. War is not meant to be pretty, and the sanitised experience of killing offered to drone pilots sometimes feels too tidy.

Manoeuvres like the raid that led to the death of Abu Ibrahim al-Hashimi al-Qurayshi are brutal, even if they are sometimes necessary. When a strategic decision involves violence, boots on the ground aid intelligence gathering, help to appease the community in which the action takes place, and reduce senseless civilian casualties.

If President Biden considered this tactic necessary to eliminate al-Qurayshi, why should the US military not take the same tack more often? The unpalatable truth is that the vast majority of targets are simply not important enough to warrant risking the safety of US military personnel, or to bother with ensuring that robust DNA evidence can be collected. A fleeting behavioural signal can be enough to condemn a ‘military-aged male’ to death. This is not good enough. If we can appreciate the benefits of the surgical mission that killed al-Qurayshi, we should seek to adopt the approach more often. Dignity for civilians and the communities in which terrorists hide should surely be taken into account alongside the raw military objective.

While we’re sadly all used to American Sniper-style hagiography by now, superhero movies are turning their gaze on emerging technologies as well as more traditional shoot-em-up conflicts. I watched Spiderman: Far From Home recently, and I was struck by how, with every fantasy fantasy under the sun at their disposal, Marvel producers felt that drones should be the key narrative device for an entertaining, self-aware, PG-certificate blockbuster.

Creating realistic combat scenarios is not something that keeps Marvel’s storytellers up at night. Spiderman: Far From Home features giant monster-smoke-tidal-waves called Elementals that wreak devastation in cities around Europe, innocent bystanders be damned. That’s OK, it’s for children. You can leave your ethics hat at the door. But Far From Home becomes more cynical, and more interesting, when Jake Gyllenhaal’s scenery-chewing villain Mysterio unveils exactly how these ‘monsters’ came to be. In fact, they aren’t supernatural forces, but cunningly designed drone swarm systems (or “illusion tech”, in Far From Home parlance) that collectively create dynamic projections of the Elementals.

A little later, in a ‘rehearsal’, we see Mysterio picking fault with the choreography of an upcoming battle. Here we realise that the damage wrought by his drones is absolutely real. Mysterio’s drones are weaponised, and although their weapons are cloaked by city-spanning wraith-like illusions, Mysterio can “double up the damage” at the flick of a switch, ramping up his drones’ destructiveness and, crucially, maximising the number of eyeballs glued to these horrific tragedies through TV and the internet.

So far, so sci-fi. But Far From Home actually does have something to say. Mysterio’s goal is to be heard; to be the purveyor of objective truth, singing from the same hymn sheet as strongmen the world over. By defeating the Elementals in battle, Mysterio can broadcast his name around the world as the most powerful superhero, way better than Iron Man, Spiderman and the rest of those do-gooders.

Mysterio recognises the deep relationship between imagery and truth. Modern news media, stacked with humanities graduates, doesn’t have the technical expertise or the integrity to exercise editorial control before broadcasting seductive, hard-hitting images. So the people who control the satellite images, in effect, control what people at home see and hear.

In the end, of course, Far From Home surrenders to the lure of superhero fantasies. When Mysterio’s illusion falls apart over Tower Bridge, he turns his drones into direct-combat weapons, hunting Spiderman at close quarters as he spaffs his way around London landmarks. Of course, the idea that drones can hunt ‘combatants’ in this way is false. Drones costing millions of dollars regularly don’t work in high winds or overly cloudy weather. No drone in active service could feasibly hunt Spiderman around SE1.

I suppose the point of all this waffle is: why should a huge superhero film use drones to power much of its plot and a lot of its character development? I think in this case the primary factor is authenticity: even casual cinema-goers know that drones are in some way unnerving and dangerous, and using drones roots the film in reality, which may help imbue showpiece scenes with more dread when you’re zooming around destroying Venice and London. But deception is also a trope here; creating illusions is central to Mysterio’s play for domination of the superhero-entertainment complex, and drones themselves rely on appearing innocuous but deadly at the same time.

Check out Spiderman: Far From Home if you want a bit of mindless fun. Oh, and one last thing: drones and the MCU are such a beautiful fit that you can find a replica 1:6 scale Mysterio’s Drones set of figurines, “beautifully painted with battle damage effects”, for £160 on Fairytale Collectables. Get ‘em while they’re hot.

The US withdrawal from Afghanistan has left a huge power vacuum that extremists are rushing to fill. It also removes one of the rhetorical pillars of the US drone programme: that unmanned weapons work alongside fighter planes and ground forces to create a proportionate but robust counter-insurgency strategy. Having no US soldiers to ‘protect’ with drones means the justification for their continued use may become even flimsier.

The withdrawal makes it logical that Afghanistan’s official designation as a “theatre of combat operations” should change. With no US troops left on the ground, it stands to reason that Afghanistan should shift to a similar state as (for example) Somalia and Yemen. These countries do harbour terrorist organisations, and the US can take military action in these countries under certain conditions, per the wide-ranging Authorization for Use of Military Force (AUMF) signed in the aftermath of 9/11 and that still remains in place nearly 20 years on, despite recent congressional votes to repeal the legislation.

But don’t let that get you thinking the US will simply dial down its use of drones in Afghanistan. Indeed, the opposite may well be true. Even while CNN reports that “the Biden administration is also still debating whether to remove the combat zone designation for Afghanistan”, the same report says that the US “will retain authority to carry out strikes against the Taliban in support of Afghan forces”.

This apparent paradox exists because of how useful drones are to military leaders. Daniel Brunstetter and Megan Braun have observed that drones “may be seen as a level of force short of war to which the principle of last resort does not apply”. This tenet, as well as the sweeping scope of the AUMF, gives US military forces – and intelligence agencies like the CIA – the latitude to use drones in countries with which the US is not at war.

Granted, it will be more expensive to use drones if the US does not have an Afghan base. Per the New York Times, writing in June:

Some officials are wary of […] drones [flying] as many as nine hours each way for a mission in Afghanistan, which would make the operations more expensive because they require more drones and fuel, and also riskier because reinforcements needed for commando raids could not arrive swiftly during a crisis.

But these worries should be put in context: drones remain far cheaper than other warfighting technologies, which are vulnerable to their own failures and inefficiencies. Take General Dynamics’ Ajax reconnaissance vehicles, touted by the UK for years as a central component of future conflicts. A UK minister said in July that the Ajax programme may have terminal problems, most notably with vehicle design issues. The FT reported that the UK has already spent £3bn on the programme.

The UK army’s solution? More drones, of course:

Lieutenant General Ralph Wooddisse, commander of the field army, said he was making contingency plans — such as using unmanned drones for reconnaissance — in case Ajax is still not in service by the middle of this decade.

The situation in Afghanistan and surrounding countries is only going to become more complex and more deadly as the US withdraws its forces. As military commanders seek to know more about territories where there is no longer a permanent military presence, drones will become even more important to Western activities in the region. Your regular reminder: drones are imprecise and unreliable. In Afghanistan alone, the Bureau of Investigative Journalism attributes a minimum of 300 and a maximum of over 900 civilian deaths to drones between the start of 2015 and today.

With everything we know about the dangers of drones, the US withdrawal should be cause for further concern rather than celebration.

Last year, Singer sought to break new ground by turning his techno-futurism into a novel, collaborating with writer and analyst August Cole to produce Burn In, “a novel of the real robotic revolution”. The premise is a common or garden sci-fi novel, only where each of the technologies described already exists in some form. Reviews have called Burn In “worryingly accurate,” but while some of the technological detail woven in to the book’s narrative is persuasively done, Burn In misrepresents some of the key conflicts at the heart of the debate into which it wades.

Lara Keegan, Burn In’s protagonist, is an FBI agent charged with training up a radical new Tactical Autonomous Mobility System, or TAMS. TAMS is a humanoid robot that is designed to partner with a human officer, but it (he?) needs to get up to speed before being deployed (and, we imagine, before many more TAMS start to roll off the production lines). Lara has to equip TAMS to respond to the innumerable grey areas created by everyday human interactions before she and the FBI can harness its next-generation capabilities appropriately and ethically.

TAMS is able to mine the trenches of the internet and surface useful information for Lara in real time – an asset to any agent. One instance sees TAMS dredging up for Lara an online conspiracy about the government’s involvement in polluting river water in a future Washington D.C. So it is jarring in the extreme to hear Lara respond with “Don’t go all Deep State on me now, TAMS”.

What? “Don’t go all Deep State on me now, TAMS.” What?? Lara is refining a revolutionary and potentially dangerous new policing technology, live in the field. Going by journalist David Rohde’s definition, ‘deep state’ means “permanent government or an institutional government […] incredibly large and powerful organizations like the FBI and the CIA and the NSA”. If you think Rohde knows what he is talking about (he does), Lara’s day job is about as ‘deep state’ as it gets.

Is it just lazy writing? Are Singer and Cole being deliberately wry in alluding to the rampant conspiracies that dog government organisations? Or are Singer and Cole deliberately painting Lara as hypocritical, demonstrating intelligence agents’ wilful blindness toward the real-world implications of their work?

The truth is, by design, difficult to pin down. Cognitive dissonance lies at the core of Western security and defence strategies. The communications agenda of Rohde’s ‘deep state’ agencies is characterised by trivia, frivolity and diversion. And it really works. What other institution’s first Instagram post would garner publicity that money can’t buy in the shape of breathless BBC articles, as we saw when the CIA made its Insta debut?

In this context, it is perhaps not so strange to see an FBI agent respond to ‘deep state’ conspiracies by immediately deflecting the conspiracy back on the accuser. Even when one’s counterpart is a robot from the future, the FBI’s attitude to any suggestion of wrongdoing, however tangential, is to deflect and deny.

In compiling an exhaustive taxonomy of live and in-development military and policing technologies, Singer and Cole presumably enjoy some access to the security state that develops and popularises these innovations. If we see a Burn In sequel, might it be worth dwelling on these dynamics a little longer, rather than skipping on to the next gadget to shoehorn in?

If you are looking for a summer read, you could probably do worse than Burn In – as long as you can put the accountability of the Western security apparatus to one side and fix your attention on the fun police robots. After all, a bit of futuristic fiction never did anyone any harm.

al-Juwaili’s assertion features in David Kirkpatrick’s coverage of developments in the messy and traumatic Libyan conflict through 2019 and 2020. Kirkpatrick’s sources include Muhammad el-Delawi, an officer in one of the many small militias that are fighting in and around Libya’s capital Tripoli, who describes the technological supremacy of the forces led by Khalifa Hifter, the commander posing the sternest challenge to Libya’s present government.

Hifter’s fighters have been furnished with “Chinese-made Wing Loong drones, purchased for $2 million each,” according to Kirkpatrick. Drones give Hifter’s forces unique battlefield advantages:

“The drone can see a fighter smoking a cigarette inside a car”, Mr el-Delawi explained.

It is true that drones can identify heat and light sources from a distance, potentially making it hard for a hypothetical fighter to smoke a cigarette without being identified. But framing el-Delawi’s comment in this way implies that drones can tell the difference between a fighter smoking a cigarette and an innocent bystander, and this is a dangerous statement for the Times to report without further analysis.

Drones do not themselves ‘see’ anything: they feed back live footage to tactical bases usually located thousands of miles away. The drone’s operator, as well as other military and legal representatives overseeing the action, are the ones doing the seeing. In real time, identifying what a flash of light might mean, or what that dark handheld shape might be, is much easier said than done. Matthew Power’s 2013 article ‘Confessions of a Drone Warrior’, in which whistleblower Brandon Bryant recounts his experience as a drone pilot, includes testimony to this effect concerning objects much larger than cigarettes:

[Bryant] was told that they were carrying rifles on their shoulders, but for all he knew, they were shepherd’s staffs.

el-Delawi is not wrong to highlight how drones impact the battlefield. It is also right for Kirkpatrick to bring el-Delawi’s perspective on the Libyan conflict to New York Times readers. But not scrutinising his comment further risks making what Ann Rogers has called “the assumption that drones save lives” even more prevalent among Western audiences. Uncritical reporting of drones’ technological advantages obscures the fact that often, drones do not live up to their reputation as ‘clean’, surgical instruments of war.

Should the New York Times apply a layer of objective analysis when certain views inadvertently risk disseminating contested or inaccurate narratives? Drones enjoy robust public support that has, at least in part, been enabled by a lack of real media scrutiny. The New York Times does publish material on the thousands of civilian deaths caused by drones. But failing to bring these perspectives into foreign correspondence means people are at risk of absorbing fragments of the issue rather than the whole picture.

“Drones contribute to civilian casualties not in spite of, but because of, their attributes”, in Rogers’ view. Media companies should seek to avoid encouraging further public complacency about what drones actually do, and the havoc they wreak on civilian lives.

Since combat drones were first deployed in the aftermath of 9/11, drones have certainly helped ease public pressure on Western leaders by reducing combatants’ exposure to battlefields. But the relationship between public opinion and military decision-making, at least as far as the use of unmanned aerial vehicles (UAVs) is concerned, has not been explored in any great depth.

As it turns out, the connection between public opinion and unmanned weapons dates back further than you might expect. Even if defence contractors are not unalloyed paragons of virtue, they can undoubtedly be prescient. In 1985, the RAND Corporation produced ‘Casualties, Public Opinion and the Vietnam War’, a report commissioned as part of its extremely deep and enduring Project AIR FORCE partnership with the US Air Force. The report’s opening pages state (italics my own):

“One of the types of contingencies in which U.S. military force may be required in the future is limited conflicts located outside the European-NATO theater […] In the wake of the Vietnam experience, it is unlikely that any U.S. administration will commit large numbers of U.S. combat personnel to sustain limited wars in the Third World without a high degree of certainty that public opinion will solidly support that commitment and continue to support it over time.”

In other words, ‘if we have to fight on behalf of non-Europeans, you can bet we’re going to do everything we can not to expend American lives along the way.’ Very cool! Regardless of the deeply cynical conditions applied to this foreign policy stance, it is interesting to see RAND making such an explicit connection between military strategy and public opinion.

The reason the RAND report was commissioned in the first place provides additional context (again, italics my own):

“Project AIR FORCE recently completed a research effort to conceptualize and evaluate the performance of advanced technology unmanned systems for employment in localized high-intensity conventional conflicts outside the European NATO theater. These systems were expressly designed to provide effective direct combat support to Third World regional allies while reducing to an absolute minimum the likelihood of substantial casualties […] to U.S. military personnel.”

That the development of lethal unmanned weapons systems was fuelled by foreign policy disasters like the Vietnam War is not a new insight.* What we learn from the Rand report, though, is that public opinion was a cornerstone of the effort to produce battlefield technologies that would keep U.S. military personnel away from danger, even while they exert force against opponents unlucky enough not to be supported by a global tech-powered warfighting infrastructure.

It would be fair to say that this strategic decision has paid off, as there appears to be little doubt that drones are currently a palatable option to the American public. A study by the Centre for a New American Security (CNAS) gave respondents ten experimental combat scenarios, ranging from low to high risk, and asked them to choose whether they would prefer a manned or unmanned military response. 

The CNAS report concluded that “in all but two of the ten experimental scenarios, the public favored unmanned aircraft over manned aircraft, confirming existing perceptions that the American public generally prefers unmanned air strikes over manned strikes.” (The exceptions were when the given military scenario involved high risk to civilian life, and when the scenario specified a low risk to air crew.)

The foreign policy goals of presidents appear to live and die by public opinion, however egregiously that opinion might conflict with established norms of international relations. In his book Kill or Capture: The War on Terror and the Soul of the Obama Presidency, journalist Daniel Klaidman describes how after a landslide victory in 2008, Barack Obama’s administration tried to kick off investigations into a number of human rights violations committed by the CIA. Even after emphasising that the investigation was more procedural than vindictive – with individual CIA agents being immune from prosecution – then-Attorney General Eric Holder still found himself hamstrung by the public’s ambivalence towards illegal interrogation techniques including waterboarding and sleep deprivation. Klaidman writes, “The [expected] outrage never materialized. The broad public just didn’t care.”

Why should that matter? Barack Obama won 365 electoral college votes in 2008, twice as many as John McCain. The Democrats controlled the Senate and the House of Representatives. Politically, the administration could have done whatever it wanted. But public opinion somehow ‘prevented’ the Justice Department from moving forward with a relatively tame probe into some actual war crimes.

Part of the reason drones are seen as viable is because the public has decided that it is worth fighting wars remotely and that a given foreign policy goal can be achieved just as expeditiously with robots rather than planes, tanks and boots on the ground. This is not the only reason drones have such broad support. But it is part of the picture.

*Andrew Cockburn’s absorbing book Kill Chain goes into detail on the unmanned aerial technologies that were being tested and deployed on Vietnamese battlefields.

The consequences of waging an illegal war are economically and politically severe, and can receive significant legislative and media scrutiny. But when wars are being debated, almost no attention is paid to whether those wars are morally just.

War is a state in which soldiers can kill other combatants and commit no crime. It is by its nature exceptional and extreme. In the words of Carl von Clausewitz, “War is an act of force [...] which theoretically can have no limits”. War is terrifying, and it is meant to be. Rightly, people have tried for centuries to regulate conduct in war, and just war theory might be the best framework we’ve got.

Just war theory has emerged from thousands of years of philosophical, religious and ethical debate. It seeks to assess how states should approach the decision to go to war (jus ad bellum) and how they conduct themselves while fighting in wars (jus in bello). 

Pacifists and pugilists are generally united in accepting just war theory as a sensible, influential code of conduct. All the same, the core principles of just war theory create some fairly significant complications. I’ll highlight two:

• The problem of legitimate authority. Just war theory says that wars must be fought by entities that have the right to wage war. In the West this basically either means a nation state or a supranational body like NATO or the UN. Unfortunately this reflects a very Western-centric worldview.
  Under just war theory, can one tribe justly declare ‘war’ on another tribe if they are grievously wronged? Was the Isis caliphate a legitimate ‘state’ that can wage war? In this second instance the Western conclusion was a resounding no. Is it hypocritical, then, for Western forces to fight against the caliphate as if it was a state, all while declaring and condemning their foe’s illegitimacy?

• The problem of discrimination. In just war theory, discrimination means distinguishing legitimate combatants from those who are not viable targets. There is a clear global consensus that soldiers should not kill non-combatants (civilians or ‘innocents’). Combatants may be identified by the wearing of uniforms; by carrying arms; and/or by intentionally aiding or sheltering other combatants.
  But war has come a long way from being a series of battles with uniformed soldiers lining up against one another in a field somewhere. In the 1960s and 1970s, the guerrilla model transformed ‘conventional’ conflict into protracted games of hunting and hiding.* More recently, the doctrine of counterinsurgency has seen Western armies fighting urban ‘insurgent’ combatants who do not wear uniforms and who fight while living among friends and family. In this environment, who exactly is fair game?

Drones place even more pressure on the tenets of just war theory. Take the legitimate authority problem I’ve just posed. It is common knowledge that the US operates drones in countries with which it is not at war, like Pakistan and Somalia. Is this just? Because of this, outside ‘official’ warzones the US delegates responsibility for drone strikes to its intelligence agencies, usually the CIA. This is not democratic: Congress does not give its consent for the CIA to kill people overseas, unlike with ‘conventional’ wars. Is this just?

Drones also test the problem of discrimination. Over a period encompassing hundreds, perhaps thousands, of Obama administration drone strikes, “the U.S. policy was to declare anyone killed in those strikes as an ‘enemy killed in action,’” per The Intercept. For years, the US failed to appropriately discriminate between combatants and non-combatants. Is this just?

It’s clear that we need to consider why wars are waged and how they are fought. There are serious problems with just war theory as a tool to do this. As Michael Walzer, the preeminent modern just war theorist, says, “It is certainly possible for a just war to be fought unjustly and for an unjust war to be fought in strict accordance with the rules.” Just war theory is imperfect, but giving it a slightly brighter spotlight when military conduct is debated in parliamentary chambers and media outlets could improve the integrity of future conflicts, whether or not autonomous weapons are involved.

—

*Gregoire Chamayou’s Manhunts is the essential read on war as hunting. Manhunts covers drones in some detail, but Chamayou’s Drone Theory offers a different and radical critique of modern militaries’ use of drones.

But the debate on the use of drones and autonomous weapons has a fatal flaw. Data on how drones are actually used are shrouded in secrecy. We do not have authoritative stats on the numbers of casualties from drone strikes, or on (for instance) the proportion of drone strikes that misidentify targets.*

The absence of trustworthy data precludes objective debate. Without data to underpin arguments, the debate has to be conducted through theory. Theory matters too, but with so much at stake, the drone debate cannot rely solely on hypotheticals.

I’ll call the camps on either side of the drone debate ‘insiders’ and ‘outsiders’.*** The ‘insider’ / ‘outsider’ binary is tied to a status quo. That status quo is the positive demeanour of military and political authorities, institutional bodies, academic contributors and the public** toward the use of drones and autonomous weapons in warfare.

By ‘insiders’, I mean those who are amenable to the prospect of drones being used in place of ‘conventional’ fighting forces (soldiers, tanks, planes) to achieve military objectives. I use ‘insider’ because the term reflects support for an incumbent position held by governments in developed nations around the world.

‘Outsiders’ sit in opposition to this stance. I define ‘outsiders’ as contributors who wish to bring about meaningful change to current drone policy, or to critique existing protocols regarding the use of unmanned weapons.

When it comes to the drone debate, two key factors threaten to unfairly skew the attitudes of policymakers, military personnel, and the public:

• The drone debate is conducted as though we are drowning in authoritative and useful data on drone strikes. In fact, we do not have anywhere near enough data to back up the assertions of insider contributors that drones in fact are more effective and more targeted than conventional air warfare.

• In drone debates, outsider voices are positioned as exceptional or radical. Insider voices occupy the ‘centre ground’, establishing what can be debated, and what should be taken for granted.

I’ll use one example that shows the difficulty of ensuring a healthy debate where different opinions are treated equally. Let’s take a volume called Opposing Perspectives on the Drone Debate, edited by a chap named Bradley Strawser. Strawser has played the role of convenor in multiple academic volumes on drones, bringing together philosophers, writers and military experts in collections of essays (Killing by Remote Control is another one to check out).

Strawser himself bridges the military and academic divide. He presently serves as an associate professor in the US Naval Postgraduate School’s Defense Analysis Department while holding a research post at Oxford University’s Institute for Ethics, Law, and Armed Conflict. He may be most well-known for his argument that if the cause is just, the use of drones should not be merely desirable but morally obligatory.

Opposing Perspectives on the Drone Debate (let’s go with OPDD) is an interesting text as it sets up a discussion between its different contributors that plays out sequentially through the volume. One chapter, ‘Even War Has Limits’, is written by Pakistani lawyer and writer Feisal Naqvi. In the context of OPDD, which spends lots of time in the abstract realms of theory, ‘Even War Has Limits’ is something of an anomaly. Naqvi attempts to deal in “physical context”, tackling the West’s use of drones in Pakistan as a Pakistani citizen and resident. He takes issue with the tone of OPDD’s other contributors, putting data – and its absence – at the centre of his argument.

In Strawser’s own first essay, ‘More Heat than Light’, he says that to weigh up issues such as drones’ long-term benefits or drawbacks, “we need good data [...] which we don’t have”. I can almost see Strawser’s shrug. Reminder: this is the same guy who has stated confidently that in an as-yet-unimagined hypothetical future situation, it is a morally questionable choice not to use drones. One might expect some evidence to support such a bold assertion as a pre-requisite, but here Strawser denies us.

Naqvi kindly goes to the trouble of pointing out the hypocrisy of this position:

Let us be clear that the reason why drone strikes are shrouded in mystery is because the United States chooses not to reveal any details. Defenders of drone strikes cannot rely on that lack of data to support their arguments because the burden of proof to justify killing people is on the United States.

Naqvi’s point is simple: to demonstrate that a military technology is morally legitimate, the aggressor needs to show that in practice, the technology in question does what it says on the tin. If the US thinks that drones kill fewer civilians, it should publish data that proves or disproves that argument.

Strawser doesn’t like this outsider perspective one bit. He replies to Naqvi in the next section of OPDD, stating that “such a polemic is unhelpful, inflammatory and wildly inaccurate; it moves us no closer to an informed, nuanced debate”. 

Although Strawser states later on that “it is not [his] intent to be the gadfly in the corner saying that [drones] could be used justly”, he makes it his responsibility to close off any arguments that don’t adhere to the strict boundaries of the debate he himself has convened. Strawser has set up a soft play centre and nobly assigned himself the responsibility of throwing out anything with sharp edges.

Insiders set and actively police the parameters of the drone debate. This is not necessarily a new phenomenon in academic circles. But this debate has serious implications for governments’ foreign policies. Strawser, for one, lectures the military leaders of the future on ethics in combat. What he thinks and publishes really matters.

One of the benefits of this being a newsletter rather than an academic journal is that I can highlight one particular argument or discussion and just focus on that. I happen to think that Strawser is arguing in good faith, and that there are sound arguments for using drones. But there is no question that inequities exist in this debate, and I think they deserve more attention.

This won’t be a regular newsletter but I will try and write more as often as I can. I definitely want to focus more on news stories and not just criticism – this was a bit dense, so if you made it all the way down here, thank you Dad for paying attention.

Be back soon.

—

*A secretive disclosure policy under the Obama administration was made even foggier when Donald Trump decided in 2019 to revoke the Obama-era policy of reporting on civilian deaths from drone strikes. Some data is available from non-governmental sources: the Bureau of Investigative Journalism is one non-governmental organisation that has tracked strikes – as best as possible – in Afghanistan, Pakistan, Libya, Somalia and Yemen, using local sources and reporting as well as official disclosures. Whether this is an effective substitute for US government data is an open question.

**The public, it appears, loves drones. This fascinating and thorough analysis by the Center for a New American Security think tank outlines the results of an in-depth survey into public attitudes towards drone strikes. Despite (or because of?) significant knowledge gaps regarding what drones do and how they work, the public almost uniformly support using unmanned rather than manned weapons in conflict scenarios.

***The boundaries between these constituencies are not set in stone. Indeed, they can be surprisingly porous: Stanley McChrystal, commander of US forces in the Iraq War, is just one prominent military figure to have raised serious questions about how the US goes about wielding military might with drones.

The question of whether autonomous weapons – particularly drones – are ‘improving’ conflict is thorny and hotly contested. A cash-strapped media occupied with many other political and social crises has failed to devote enough space to drones and other autonomous weapons.

My Master’s thesis explored the idea of drones on film, and how films about drones may serve political purposes for the countries using them. Working for several technology startups has also given me some insight into how technology can create change – destructive and redemptive – at speed.

Sign up to receive future newsletters:

Subscribe now

And in the meantime, do tell your friends!

*The name of this newsletter is a bastardised allusion to the Campaign to Stop Killer Robots, a coalition of NGOs aiming to prevent the deployment of fully autonomous weapons in conflicts. Read more about the campaign here: stopkillerrobots.org. For more information on the use of drones around the world, check out the work of Bureau of Investigative Journalism and Bellingcat, to name but two vital publications.