Drone networks

Posted on February 22, 2015 by Derek Gregory

Three contributions to the debates over drones and military violence. First, my friends at the Bard Center for the Study of the Drone have published Dan Gettinger‘s essay on ‘Drone Geography: mapping a system of intelligence‘. It’s a superb sketch of the intelligence network in which the US Air Force’s drones are embedded (you can read my complementary take on ‘Drone geographies’ under the DOWNLOADS tab). Let me add just one map to the illustrations that stud his essay. It’s taken from the Air Force’s RPA Vector report for 2013-28, published last February, and it shows the architecture of remote split operations within and beyond the United States. It’s helpful (I hope) because it shows how the Ground Control Stations in the continental United States feed in to the Distributed Common Ground System that provides image analysis and exploitation (shown in the second map, which appears in a different form in Dan’s essay). I’m having these two maps combined, and I’ll post the result when it’s finished.

Dan is right to emphasise the significance of satellite communications; much of the discussion of later modern war and its derivatives has focused on satellite imagery, and I’ve discussed some of its complications in previous posts, but satellite communications materially shape the geography of remote operations. The Pentagon has become extraordinarily reliant on commercial providers (to such an extent that Obama’s ‘pivot to the Pacific’ may well be affected), and limitations of bandwidth have required full-motion video streams from Predators and Reapers (which are bandwidth hogs) to be compressed and image quality to be degraded. Steve Graham and I are currently working on a joint essay about these issues.

One caveat: this is not the only network in which US remote operations are embedded. In my essay on ‘Dirty Dancing’ (now racing towards the finish line) I argue that the CIA-directed program of targeted killing in the Federally Administered Tribal Areas of Pakistan cannot be severed from the multiple ways in which the FATA have been configured as both borderlands and battlefields and, in particular, from the cascade of military operations that have rendered the FATA as a space of exception (in something both more and less than Agamben’s sense of the term). Here I’ve learned much from an excellent essay by Elizabeth Cullen Dunn and Jason Cons, ‘Aleatory sovereignty and the rule of sensitive spaces’, Antipode 46 (1) (2014) 92-109). They complicate the claim that spaces of exception always derive from a single locus of sovereign power (or ‘the sovereign decision’). Instead, they suggest that borderlands are ‘contested spaces’ where ‘competing’ powers ‘collide’. In the FATA multiple powers have been involved in the administration of military violence, but on occasion – and crucially – they have done so in concert and their watchword has been a qualified and covert collaboration. In particular, the FATA have been marked by a long and chequered gavotte between the militaries and intelligence services of the United States and Pakistan which, since the 1980s, has consistently put at risk the lives of the people of the borderlands. And in my essay on ‘Angry Eyes’ (next on my screen) I argue that the US military’s major use of Predators and Reapers in Afghanistan – orchestrating strikes by conventional aircraft and providing close air support to ‘troops in contact’ – depends on communication networks with ground troops in theatre, and that this dispersed geography of militarised vision introduces major uncertainties into the supposedly ‘precise’ targeting process.

Second: Elliott Prasse-Freeman has an extended review of the English translation of Grégoire Chamayou‘s Theory of the drone – called ‘Droning On‘ – over at the New Inquiry (you can access my own commentaries on the French edition here: scroll down). His central criticism is this:

While his title promises theory, we instead are treated to a digression on the military and social ethics of attacks from the air, in which Chamayou asks without irony, “can counterinsurgency rise to the level of an aero-policy without losing its soul?” What offends Chamayou is the “elimination, already rampant but here absolutely radicalized, of any immediate relation of reciprocity” in warfare. This, we are told, is the problem.

Promised a theory of the drone, how do we arrive at a theory of the noble soldier?…

And so, dispatching with the dream of the drone … Chamayou assumes the concerns not of the brutalized but of military leaders and soldiers.

He continues in terms that resonate with my argument in ‘Dirty Dancing’:

By combining knowing (intelligence, surveillance, reconnaissance), sighting (targeting in movement and in the moment), and eliminating (“putting warheads on foreheads”), the drone constitutes an assemblage of force (as drone-theorist Derek Gregory puts it) that promises a revolution in control and allows the US war apparatus to imagine space and politics in new ways. Because the body of the accused can ostensibly be precisely seen, it can be seen as itself carving out a body-sized exception to state sovereignty over the territory on which that body moves. In this way, eliminating the body does not constitute an assault on the territory of the state, as these bodies are presented as ontologically (and hence quasi-legally) disconnected from that territory. Geographer Stuart Elden in Terror and Territory (2009) points out the significant overlap between who are labeled ‘terrorists’ and movements fighting for their own political spaces – which hence necessarily violate extant states’ ‘territories’ (and hence the entire international order of states): to violate territory is to terrorize. The US is hence remarkably concerned in its arrogation of a position of supra-sovereignty to ensure that it overlaps with ‘classic’ state sovereignty, and by no means violates the norm of territorial integrity (well-defined borders): by harboring or potentially harboring unacceptable transnational desires, the militant uproots himself, and risks being plucked out and vaporized in open space that belongs only to him. The exception to sovereignty provides the drone the opportunity to extend this exception into temporal indefiniteness: wars are not declared, aggressions are not announced—the fleet, fusing police and military functions, merely watches and strikes, constantly pruning the ground of human weeds.

In ‘Dirty Dancing’ I’m trying to prise apart – analytically, at least – the space of exception, conceived as one in which a particular group of people is knowingly and deliberately exposed to death through the political-juridical removal of legal protections and affordances that would otherwise be available to them, and territory conceived (as Stuart suggests) as a political-juridical technology, a series of calculative practices that seeks to calibrate and register a claim over bodies-in-spaces. That’s why Dan Gettinger’s essay is so timely too, and why I’ve been thinking about the FATA as a performance of what Rob Kitchin and Martin Dodge call ‘code/space’, why I’ve been working my way through the files released by Edward Snowden, and why I’ve been thinking so much about Louise Amoore‘s superb critique of The politics of possibility: risk and security beyond probability (2013).

Although Louise doesn’t address drone strikes directly, her arguments are full of vital insights into the networks that are mobilised through them. ‘The sovereign strike is always something more, something in excess of a single flash of decision’, she insists, and when she writes that —

those at risk (which is to say those who are to be put at risk by virtue of their inferred riskiness) are ‘not strictly “included by means of their own exclusion”, as Agamben frames the exception, they are more accurately included by means of a dividuated and mobile drawing of risk fault lines’

— it’s a very short journey back to Grégoire Chamayou‘s reflections on the strange (in)dividual whose ‘schematic body’ emerges on the targeting screen of the Predator or Reaper. Louise writes of ‘the appearance of an emergent subject’, which is a wonderfully resonant way of capturing the performative practices through which targets are produced: ‘pixelated people’, she calls them, that emerge on screens scanning databanks but which also appear in the crosshairs…

And finally, Corporate Watch has just published a report by Therezia Cooper and Tom Anderson, Gaza: life beneath the drones. This brings together a series of interviews conducted in 2012 – when ‘drones killed more people in Gaza than any other aircraft’ – that were first published in serial form in 2014. The report includes a tabulation of deaths from Israeli military action in Gaza and those killed directly by drones (2000-2014) and a profile of some of the companies involved in Israel’s military-industrial complex.

Security archipelagos

Posted on March 20, 2014 by Derek Gregory

Three short contributions that have caught my eye raise a series of interesting questions about contemporary ‘security archipelagos’ (in multiple sense of the term, hence the plural).

The term itself comes from Paul Amar, and Austin Zeiderman has a short but interesting review of his The Security Archipelago: Human-Security States, Sexuality Politics, and the End of Neoliberalism (Duke, 2013) over at Public Books (Public Culture‘s public site):

‘Amar asserts that we need an analytical framework focused on the rise of human security—a governance regime that “aim[s] to protect, rescue, and secure certain idealized forms of humanity.” This new regime is gradually replacing neoliberalism, Amar contends, “as the hegemonic project of global governance and of state administration.” This shift is evident in how security is now justified and pursued by states. The antagonistic relationship between security and human rights that characterized the “neoliberal market states” of the late 20th century is no longer so evident. The repressive security strategies that underpinned earlier development paradigms have been succeeded by the “promise to reconcile human rights and national security interests” in the interest of economic prosperity. Progressive and conservative security doctrines now agree on the imperative to “humanize” (or “humanitarianize”) both state and parastatal security apparatuses. The result, Amar argues, is what he calls the “human-security state”: a globally emergent governance regime with “consistent character and political profile.” From Latin America to the Middle East, political legitimacy is increasingly based on securing humanity against a range of malicious forces….

If the megacities of the Global South are indeed “laboratories” in which new logics and techniques of global governance are being created, it is up to other researchers to fill out and develop further Amar’s concept of the “security archipelago.” Though his study provides both the theoretical rationale and the analytical tools with which to do so, it may be worth questioning whether the “human” is necessarily central to emerging security regimes. For along with human security apparatuses and the human actors struggling to articulate progressive alternatives, a host of non-humans—drones, border fences, hurricanes—are actively producing the security landscape of the future.’

Secondly, I’ve been thinking about the ways in which the work of these ‘laboratories’ often relies on non-state, which is to say corporate, commercial sites (this isn’t news to Paul, of course, even if he wants to challenge our ideas about neoliberalism). We surely know that the traditional concept of the military-industrial complex now needs wholesale revision, and I’ve noted before the timely and important essay by Jeremy Crampton, Sue Roberts and Ate Poorthuis on ‘The new political economy of geospatial intelligence‘ in the Annals of the Association of American Geographers 104 (1) (2014) (to which I plan to return in a later post). The latest MIT Technology Review has a short but suggestive essay by Antonio Regalado, ‘Spinoffs from Spyland’, which describes some of the pathways through which the National Security Agency commercializes (and thus potentially subcontracts and, in some cases, even subverts) its surveillance technology:

In 2011, the NSA released 200,000 lines of code to the Apache Foundation. When Atlas Venture’s Lynch read about that, he jumped—here was a technology already developed, proven to work on tens of terabytes of data, and with security features sorely needed by heavily regulated health-care and banking customers. When Fuchs’s NSA team got cold feet about leaving, says Lynch, “I said ‘Either you do it, or I’ll find five kids from MIT to do it and they’ll steal your thunder.’”

Eventually, Fuchs and several others left the NSA, and now their company [Sqrrl] is part of a land grab in big data, where several companies, like Splunk, Palantir, and Cloudera, have quickly become worth a billion dollars or more.

Over the summer, when debate broke out over NSA surveillance of Americans and others, Sqrrl tried to keep a low profile. But since then, it has found that its connection to the $10-billion-a-year spy agency is a boost, says Ely Kahn, Sqrrl’s head of business development and a cofounder. “Large companies want enterprise-scale technology. They want the same technology the NSA has,” he says.

And finally, before we rush to radicalise and globalise Foucault’s critique of the Panopticon, it’s worth reading my friend Gaston Gordillo‘s cautionary note – prompted by the search for missing Malaysian Airlines Flight 370 – on ‘The Opaque Planet’:

The fascination with, and fetishization of, technologies of global location and surveillance often makes us forget that, for all their sophistication, we live on a planet riddled with opaque zones that will always erode the power of human-made systems of orientation, for the simple fact that no such system (contrary to what the NSA seems to believe) will ever manage to create an all-seeing God. This opacity is intrinsic to the textured, three-dimensional materiality of the surface of the planet, and is especially marked in the liquid vastness of the ocean.

Phil Steinberg has already commented on the geopolitics of the search, but Gaston draws out attention to the gaps in the surveillance capabilities of states, and here too the geopolitical meshes (and sometimes jibes against) the geoeconomic, as described in this report from Reuters:

Analysts say the gaps in Southeast Asia’s air defenses are likely to be mirrored in other parts of the developing world, and may be much greater in areas with considerably lower geopolitical tensions.

“Several nations will be embarrassed by how easy it is to trespass their airspace,” said Air Vice Marshal Michael Harwood, a retired British Royal Air Force pilot and ex-defense attache to Washington DC. “Too many movies and Predator (unmanned military drone) feeds from Afghanistan have suckered people into thinking we know everything and see everything. You get what you pay for. And the world, by and large, does not pay.”

Killing the SIMs

Posted on February 10, 2014 by Derek Gregory

Hard on the heels of yesterday’s post about Metadata+ – and my earlier, necessarily impressionistic discussion of the National Security Agency, drone strikes in Pakistan and what I called ‘the everyware war’ – comes a report from Jeremy Scahill and Glenn Greenwald at The Intercept that details the NSA’s role in the US’s targeted killing programme.

The programme is carried out by multiple means and multiple agencies (including the US military), but the main source for their report is a former drone operator from Joint Special Operations Command (JSOC) – and readers of Jeremy’s Dirty Wars will know that it is JSOC rather than the CIA that is his main critical target. We’ve known about the NSA’s Geolocation Cell (‘GeoCell’) for some time now. Last summer Dana Priest outlined the history of collaboration between the National Geospatial Intelligence Agency and the NSA, which she traced back to the months following 9/11, and the determination to use cell phones as ‘targeting beacons’:

The cell opened up chat rooms with military and CIA officers in Afghanistan — and, eventually, Iraq — who were directing operations there. Together they aimed the NSA’s many sensors toward individual targets while tactical units aimed their weaponry against them.

A motto quickly caught on at Geo Cell: “We Track ’Em, You Whack ’Em.”

But, as Scahill and Greenwald show, the process remains far from reliable. Their informant confirms that strikes are often carried out on the basis of metadata analysis and cellphone tracking (‘SIGINT’) rather than human intelligence (agents on the ground), and he admits that it is a hit or miss affair, and literally so: ‘high value targets’ have been killed, but so too have innocent people.

[T]he geolocation cells at the NSA that run the tracking program – known as Geo Cell – sometimes facilitate strikes without knowing whether the individual in possession of a tracked cell phone or SIM card is in fact the intended target of the strike.

“Once the bomb lands or a night raid happens, you know that phone is there,” he says. “But we don’t know who’s behind it, who’s holding it. It’s of course assumed that the phone belongs to a human being who is nefarious and considered an ‘unlawful enemy combatant.’ This is where it gets very shady.”

There are several lists of so-called High Value Targets, but the operational problem is, as he concedes, ‘We’re not going after people – we’re going after their phones.’

[T]racking people by metadata and then killing them by SIM card is inherently flawed. The NSA “will develop a pattern,” he says, “where they understand that this is what this person’s voice sounds like, this is who his friends are, this is who his commander is, this is who his subordinates are. And they put them into a matrix. But it’s not always correct. There’s a lot of human error in that.”

The programme relies on more than what we have come to think of as standard (though still shocking) NSA intercepts of communications. For more on the NSA’s global tracking of cellphones, see Barton Gellman and Ashkan Soltani here, from which I’ve taken the image below.

As the graphic shows, the standard method of ‘co-location analytics’ tracks not only targets but also their associates (‘co-travellers’) using cell towers (which have in turn become key targets for insurgent attack in Afghanistan partly for this reason).

For clandestine operations in Afghanistan, Pakistan, Yemen, Somalia and elsewhere the NSA’s geolocation system codenamed ‘GILGAMESH’ (king of the city-state of Uruk in Mesopotamia: where do they get these names from?) equips Predators assigned to JSOC with ‘virtual base-tower transceivers’ that can track and locate a SIM card to within 30 feet; the information is then used in near-real time to activate an air strike or a night raid through a network that has ‘cued and compressed numerous kill chains’.

The CIA uses another system, codenamed SHENANIGANS (no mystery about that one), which collects data ‘from any wireless routers, computers, smart phones or other electronic devices that are within range.’ Last fall Greg Miller, Julie Tate and Barton Gellman described how the system was used in the targeted killing of Hassan Ghul in Mir Ali in Waziristan in October 2012:

In the search for targets, the NSA has draped a surveillance blanket over dozens of square miles of northwest Pakistan. In Ghul’s case, the agency deployed an arsenal of cyber-espionage tools, secretly seizing control of laptops, siphoning audio files and other messages, and tracking radio transmissions to determine where Ghul might “bed down.”

The e-mail from Ghul’s wife “about her current living conditions” contained enough detail to confirm the coordinates of that household, according to a document summarizing the mission. “This information enabled a capture/kill operation against an individual believed to be Hassan Ghul on October 1,” it said.

The file is part of a collection of records in the Snowden trove that make clear that the drone campaign — often depicted as the CIA’s exclusive domain — relies heavily on the NSA’s ability to vacuum up enormous quantities of e-mail, phone calls and other fragments of signals intelligence, or SIGINT.

To handle the expanding workload, the NSA created a secret unit known as the Counter-Terrorism Mission Aligned Cell, or CT MAC, to concentrate the agency’s vast resources on hard-to-find terrorism targets. The unit spent a year tracking Ghul and his courier network, tunneling into an array of systems and devices, before he was killed. Without those penetrations, the document concluded, “this opportunity would not have been possible.”

The two systems are agency-specific but they often work in concert: Scahill and Greenwald note one joint operation in 2012, ‘VICTORYDANCE’, that ‘mapped the WiFi fingerprint of nearly every town in Yemen.’

The documents that they draw upon are from various dates, extending back to 2005, so that these are a series of snapshots of an uneven and developing set of systems. Still, the concerns about accuracy are persistent, and although they don’t cite her report Kate Clark‘s forensic examination of the attempted killing of the Taliban shadow governor of Takhar province in Afghanistan in September 2010 provides an exceptionally detailed account of the dangers of relying on metadata and cellphone tracking. Here is what I said about this in my discussion of Targeted Killing and Signature Strikes:

On 2 September 2010 ISAF announced that a ‘precision air strike’ earlier that morning had killed [Muhammad Amin] and ‘nine other militants’. The target had been under persistent surveillance from remote platforms – what Petraeus later called ‘days and days of the unblinking eye’ – until two strike aircraft repeatedly bombed the convoy in which he was travelling. Two attack helicopters were then ‘authorized to re-engage’ the survivors. The victim was not the designated target, however, but Zabet Amanullah, the election agent for a parliamentary candidate; nine other campaign workers died with him. Clark’s painstaking analysis clearly shows that one man had been mistaken for the other, which she attributed to an over-reliance on ‘technical data’ – on remote signatures. Special Forces had concentrated on tracking cell phone usage and constructing social networks. ‘We were not tracking the names,’ she was told, ‘we were targeting the telephones.’

The Takhar case has been on my mind again because, as Kate revealed in a new report last week, the agencies involved in targeted killing reach beyond the NSA – and beyond the USA.

The involvement of Britain’s GCHQ and other ‘Five Eyes’ sites in global surveillance has become well-known in recent months – and see Nikolas Kozloff for a short reflection on its ‘geopolitics of racism‘ (though he ignores the willingness to spy on other European states and allies) – but Kate’s new report is significant for what it reveals about the complicity of civilian police agencies. In August 2013 Habib Rahman, who lost two brothers, two uncles and his father-in-law in the Takhar attack, petitioned for a judicial review of the involvement of Britain’s Ministry of Defence and the (civilian) Serious Organised Crime Agency (SOCA) – whose brief includes the drug trade from Afghanistan – in the compilation of the Joint Prioritised Effects List (one of the lists of priority targets used by the Pentagon).

SOCA is part of the Joint Inter-Agency Afghanistan Task Force (JIATF), a military-police task force based at Kandahar AFB that also includes the US Drug Enforcement Agency, the US military and the British military. In a US Senate report on the nexus between drug trafficking and insurgency, ‘Afghanistan’s Narco War‘, a SOCA investigator described the proposed collaboration as ‘a critical opportunity to blend military and law enforcement expertise’ (p.15). At the time (August 2009) the JIATF was awaiting formal approval from Washington and London, but informal collaboration was already under way, and the SOCA officer was enthusiastic about its consolidation:

‘In the past, the military would have hit and evidence would not have been collected,’ he explained. ‘Now, with law enforcement present, we are seizing the ledgers and other information to develop an intelligence profile of the networks and the drug kingpins.’ An American military officer with the project was blunter, telling the committee staff, ‘Our long-term approach is to identify the regional drug figures and corrupt government officials and persuade them to choose legitimacy or remove them from the battlefield.’

Rahman’s lawyers argued:

As a civilian policing organisation SOCA has no legitimate or lawful role to play in the compilation or administration of this Kill List – it has no authority to be involved in military operations and the killing of individuals. The courts must urgently review whether the SOCA’s and indeed the UK’s role in the compilation, review and execution of this list if unlawful. Incidents like that affecting our client must be properly investigated.

In response, SOCA acknowledged making its intelligence available to the military, but added these provisos:

SOCA places restrictions on the dissemination of intelligence to the ISAF. Intelligence which is disseminated by SOCA is required to include handling conditions which require the express approval of the originator if it is proposed to use the material for military targeting purposes. If the mission is to arrest, with a view to criminal investigation and potential prosecution, SOCA would ordinarily be prepared to provide and/or allow the use of its intelligence. On the other hand, if the primary option for a mission is to use lethal force, SOCA would not provide intelligence or allow the use of its intelligence in support of such a mission, save potentially where the individual who is the target of the operation poses a significant and immediate threat to he lives of others (and so such disclosure would be for the purpose of preventing a criminal offence).

The judge dismissed Rahman’s claim. But read that last sentence again. These missions are almost always designated as ‘kill or capture’, and while in practice ‘kill’ is often preferred over ‘capture’, how likely is it that the ‘primary option for a mission’ would be advertised as a targeted killing before the intelligence were made available? An undated US Justice Department memo setting out conditions for targeted killing included the stipulation that where ‘capture is infeasible’ the US would continue ‘to monitor whether capture becomes feasible’, which turns the mission (and SOCA’s proviso) into a moving target in more ways than one. Once intelligence has been incorporated into the matrix it can hardly be withdrawn or embargoed.

And that final clause about pre-emptive action opens up a discretionary space that the Obama administration has already sought to enlarge with its demands for an ‘elongated’ concept of imminence in which the ‘immediacy’ of the threat is subject to prolongation to the point of contortion.

Rahman is not the first victim to have been rebuffed by British courts. Last month the Court of Appeal dismissed Noor Khan‘s case against the British Government over the role of GCHQ in providing ‘locational intelligence’ for a CIA-directed drone strike in March 2010 in Miranshah in North Waziristan that killed his father. The ruling accepted that the case put forward by Noor Khan’s barrister – that the provision of such intelligence risked contravening the Serious Crimes Act (2007) – was ‘attractive’ and ‘persuasive’, but concluded that the review could not proceed because the claims

‘involve serious criticisms of the acts of a foreign state. It is only in certain established circumstances that our courts will exceptionally sit in judgment of such acts. There are no such exceptional circumstances here.’

The state of exception, it turns out, is not so exceptional… To be continued.