What Kind of Drones Arms Race Is Coming?

New York Times national security correspondent Scott Shane has an opinion piece in today’s Sunday Times predicting an “arms race” in military drones. The methodology essentially looks at the US as the leader, followed by Israel — countries that have built, deployed and used drones in both surveillance and as weapons platforms. It then looks at the list of other countries that are following fast in US footsteps to both build and deploy, as well as purchase or sell the technology — noting, correctly, that the list is a long one, starting with China. The predicament is put this way:

Eventually, the United States will face a military adversary or terrorist group armed with drones, military analysts say. But what the short-run hazard experts foresee is not an attack on the United States, which faces no enemies with significant combat drone capabilities, but the political and legal challenges posed when another country follows the American example. The Bush administration, and even more aggressively the Obama administration, embraced an extraordinary principle: that the United States can send this robotic weapon over borders to kill perceived enemies, even American citizens, who are viewed as a threat.

“Is this the world we want to live in?” asks Micah Zenko, a fellow at the Council on Foreign Relations. “Because we’re creating it.”

By asserting that “we’re” creating it, this is a claim that there is an arms race among states over military drones, and that it is a consequence of the US creating the technology and deploying it — and then, beyond the technology, changing the normative legal and moral rules in the international community about using it across borders. In effect, the combination of those two, technological and normative, forces other countries in strategic competition with the US to follow suit.

It sounds like it must be true. But is it? There are a number of reasons to doubt that moves by other countries are an arms race in the sense that the US “created” it or could have stopped it, or that something different would have happened had the US not pursued the technology or not used it in the ways it has against non-state terrorist actors. Here are a couple of quick reasons why I don’t find this thesis very persuasive, and what I think the real “arms race” surrounding drones will be.

Unmanned aerial vehicles have clearly got a big push from the US military in the way of research, development, and deployment. But the reality today is that the technology will transform civil aviation, in many of the same ways and for the same reasons that another robotic technology, driverless cars (which Google is busily plying up and down the streets of San Francisco, but which started as a DARPA project), will eventually have an important place in ordinary ground transport. UAVs will eventually move into many roles in ordinary aviation, because it is cheaper, relatively safer, more reliable — and it will eventually include cargo planes, crop dusting, border patrol, forest fire patrols, and many other tasks. There is a reason for this — the avionics involved are simply not so complicated as to be beyond the abilities of many, many states. Military applications will carry drones many different directions, from next-generation unmanned fighter aircraft able to operate against other craft at much higher G stresses to tiny surveillance drones. But the flying-around technology for aircraft that are generally sizes flown today is not that difficult, and any substantial state that feels like developing them will be able to do so.

But the point is that this was happening anyway, and the technology was already available. The US might have been first, but it hasn’t sparked an arms race in any sense that absent the US push, no one would have done this. That’s just a fantasy reading of where the technology in general aviation was already going; Zenko’s ‘original sin’ attribution of this to the US opening Pandora’s box is not a credible understanding of the development and applications of the technology. Had the US not moved on this, the result would have been a US playing catch-up to someone else. For that matter, the off-the-shelf technology for small, hobbyist UAVs is simple enough and available enough that terrorists will eventually try to do their own amateur version, putting some kind of bomb on it.

Moving on from the avionics, weaponizing the craft is also not difficult. The US stuck an anti-tank missile on a Predator; this is also not rocket science. Many states can build drones, many states can operate them, and crudely weaponizing them is also not rocket science. The US didn’t spark an arms race; this would occur to any state with a drone. To the extent that there is real development here, it lies in the development of specialized weapons that enable vastly more discriminating targeting. The details are sketchy, but there are indications from DangerRoom and other observers (including some comments from military officials off the record) that US military budgets include amounts for much smaller missiles designed not as anti-tank weapons, but to penetrate and kill persons inside a car without blowing it to bits, for example. This is genuinely harder to do — but still not all that difficult for a major state, whether leading NATO states, China, Russia, or India. The question is whether it would be a bad thing to have states competing to come up with weapons technologies that are … more discriminating.

The real place where states divide in their technological abilities with respect to drones is not about avionics or weapons — but the third, crucial conceptual element of a drone: sensor capabilities. Those are capabilities in both hardware, the kinds of signals that can be taken in from the real world (video, radar, infrared, etc.), and software, the kinds of analytic integration that can be achieved, coupled with other streams from outside the drone, such as general telecomm monitoring. That is an area in which the US has a significant edge — the Chinese, Russians, India, or for that matter, Germany or France, could do it if desired, though everyone will find it easier to steal it and reverse engineer.

Worth noting, as well, something that is rarely noted: today’s sensor technologies work well over deserts and largely bare mountains. They have not been developed for seeing through forest or jungle cover. Moreover, none of this takes account of a central reason why the US is successful with drones in Pakistan — the investment over years in an on-the-ground intelligence network that permits targeting in the first place: the focus on drones as technology in the Shane piece doesn’t not pay enough attention to this crucial, non-technological element.  The recent Reuters piece on the ground-level intelligence piece makes it clear that purely focusing on the technology gives a profoundly misleading picture of drones and their capabilities.

But again, let’s ask, why does a state want more advanced sensor technology? Yes, better sensor technology does provide greater intelligence in targeting. And China will certainly want that, because it will find that it has reasons for wanting to engage in targeted assassination against non-state actors or, for that matter, state targets. But an awful lot of countries that want drones are not really that picky, because they don’t fundamentally care much about greater discrimination in targeting; they are not that worried about indiscriminate attack, either because, as in the case of Hamas targeting Israel, indiscriminate attack is the point or because they don’t care about collateral damage so long as the target is destroyed. So why invest in greater discrimination achieved through expensive and perhaps inaccessible sensor technologies? If there were going to be an arms race in drones, it would take place here, where the technology is not already widely available — but the parties who want it are limited.

The real arms race in drones will take place, not around drones themselves, but in counters to drones. Drones are effective, in their current form, against non-state actors, terrorists, low tech insurgents, because they have no air-defense systems. The game changer in Afghanistan against Soviet helicopters was American-supplied Stinger missiles. When some state designs and starts handing out some form of air-defense system to non-state actors, then the arms race in drones actually begins. It begins with re-designs of slow, noisy surveillance craft designed to operate at high altitude over long periods of time. It takes into account one of the most likely forms of counter to drones — viruses or other malware that interferes with the communications links that control the drones — shades of the malware discovered in drone systems recently.

It is indeed likely that the future will see more instances of uses of force at a much smaller, often less attributable, more discrete level than conventional war. Those uses will be most easily undertaken against non-state actors, rather than states, though the difference is likely to erode.  The idea that it would not have occurred to China or Russia that drones could be used to target non-state actors across borders in safe havens, or that they would not do so because the United States had not done so is far-fetched.  That is so not least because the United States has long held that it, or other states threatened by terrorist non-state actors in safe havens across sovereign borders, can be targeted if the sovereign is unable or unwilling to deal with them.  There’s nothing new in this as a US view of international law; it goes back decades, and the US has not thought it some special rule benefiting the US alone.  So the idea that the US has somehow developed this technology and then changed the rules regarding cross-border attack on terrorists is just wrong; the US has believed this for a long time and thinks it is legally and morally right.

Then there a further idea that drones make it “too easy” to reach across borders and that is the difference today; a long-standing legal doctrine suddenly made far too powerful by reason of new technology.  I am not convinced.  That drones — precisely because they are accepted as both more sparing of civilians and more sparing of one’s own forces — makes it “too easy” to use force, reduces the disincentive against using force, has proven irresistible to many as a criticism of drones and targeted killing. I address some of the questions in this draft article. Still, one consideration is simply that the number of “resorts to force” is not enough to damn drones and targeted killing. One must also consider the intensity of the fighting that ensues by comparison to conventional war, as well as the question of whether they increase or diminish the damage that might otherwise arise from conventional wars that take place in lieu of these more discrete uses of force.

The moral or legal case against drones and targeted killing, because their jus in bello virtues supposedly increase the propensity to use force, is not obvious, at least not to me. But in any case, this is almost certainly the future direction of uses of force. Shane is falling into a trap here not dissimilar to the advocates who believed that they could forestall the militarization of airplanes and aerial bombardment in the early 20th century. Today’s targeted killing technologies aboard drones is finally providing a technological advance — not a solution, certainly, but an advance — on the problem of indiscriminate aerial bombardment that begin with airplanes in the 20th century. It’s not just that this technology is coming, in other words — it is that, seen over the whole course of the history of military aviation, this is a good thing, not a bad thing. It’s a pity that the Times article doesn’t see this — and given what Scott Shane and Mark Mazzetti wrote in their last, outstanding piece on drones (correctly observing that drones are about non-state actors, not interstate conflict, because drones are simply too vulnerable today), surprising to me.

There are indeed important ways in which state practice needs to evolve to deal with the legal and moral implications of new technologies and their strategic implications. One is the development of a sort of “state practice” of “intelligence-driven, discrete uses of force” — a clumsy term for what amounts to evolving “covert,” but often not covert, action. That’s where the US needs to lead the way — in the development of state practice to assert that even these emerging forms of using force are, in the first place, subject to the basic customary obligations for any use of force: necessity, distinction, proportionality. I have said a couple of times, mostly to amusement or incredulity, that the beginning that the United States under both the Bush and Obama administrations is making toward developing state practice that can be asserted as standards for the United States and other actors might well turn out to be Harold Koh’s most important contribution to international law as legal adviser to the State Department. The Barron-Lederman memo on the targeting of Al-Awlaki is an important domestic law homologue to that.

But return to the Scott Shane article. It is simply implausible to think that countries would not have been developing UAVs for military uses, just as they are being developed and deployed for civilian uses. The US might have been first, but this is where civilian aviation, and a lot of other robotic technologies, have been going even if only now becoming visible to the broader public. The deployment of weaponized drones by the US is even less the morality fable that Zenko suggests and Shane endorses as a moralizing rebuke to the United States in the Times piece. The real struggle begins over counter-technologies to drones, and counters to the counters — and that, ad infinitum.

(Update:  A friend whose views I always take seriously tells me that I am putting too much weight on “we’re” and overinterpreting Zenko’s remark.  That’s possible.  However, in that case, simply treat my remarks as against the view that the United States is the prime mover here, by inventing a technology and then altering the prior rules to suit its use.  If I misinterpret Zenko or, for that matter, Shane here, my apologies, but I think the general position, against which I’m arguing, is important if not ultimately persuasive.)

(Further update: The other unstated premise underlying the whole opinion piece is a studiously neutral moral relativism signaled by that otherwise unexamined phrase “perceived enemies.” Does it matter if they are not merely our “perceived” but are our actual enemies?  Irrespective of what one might be entitled to do to them, is it so very difficult to conclude, even in the New York Times, that Anwar al-Awlaki was, in objective terms, our enemy?)