In one respect, Lord Butler’s report was odd. As we have seen, GCHQ produces the majority of Britain’s intelligence. Astonishingly, in the 216 pages of the report ‘GCHQ’ appears only once, and that is in the list of abbreviations. It seems that Cheltenham was referred to in the text of an earlier draft, but in the published version all discussion of it has mysteriously disappeared. In fact Peter Freeman, a veteran GCHQ officer who served as the link between Cheltenham and the inquiry team, persuaded Butler to remove all direct references to GCHQ from the report.76 The Butler inquiry is also an example of the masterly drawing of remits. Although it was not permitted to look at the link between intelligence and policy-making over Iraq, which is what everyone wanted to know about, curiously, it was allowed to make meandering historical digressions into intelligence on other WMD episodes as far back as 1990. These included work against the AQ Khan proliferation network in Pakistan and the winding up of President Gaddafi’s nuclear programme in Libya. The case studies were selective: Butler went far enough back to find some successes, but not far enough to find other failures.77
Butler’s inquiry into British intelligence, and parallel inquiries in the United States, underlined the strange new climate the intelligence agencies were working in. Although not exactly transparent, they were under the spotlight of investigation as never before. Trust in them had been badly corroded, and many thought it would take a decade to repair the damage. International teams of journalists, aided by dissident officials, were ripping the lid off each sensitive story in short order. Moreover, the British public was now highly suspicious about the increasing levels of surveillance at home and abroad. Yet simultaneously, the upsurge in international terrorism had led governments to call for a massive increase in secret intelligence activity of every kind. Understandably, no politicians wanted another 9/11 to happen on their watch. The balance between liberty and security was already a hot issue, and during 2005 it would become even hotter.78
26
From Bletchley Park to a Brave New World?
These new proposals suggest an intention to capture anything and everything…
UK internet service providers commenting on government plans, August 20091
At 8.50 on the morning of Thursday, 7 July 2005, four suicide bombers attacked London. Following the classic al Qaeda approach of multiple strikes, they bombed three tube trains within the space of a minute, and an hour later attacked a bus in Tavistock Square. Fifty-two people died and over seven hundred were injured in what proved to be the most deadly attack on the capital since the Second World War. Tony Blair was away at the G-8 Summit at Gleneagles in Scotland, as indeed were numerous specialist police units from London, sent there to guard the visiting world leaders from marauding anti-globalisation protesters. In the immediate wake of the attacks, George Bush and Italy’s Prime Minister Silvio Berlusconi argued over whether the G-8 leaders should come to London to show solidarity. In the end, Blair left his international guests in Scotland and flew back to London to join his Cabinet colleagues in the increasingly familiar surroundings of ‘Cobra’, deep beneath Whitehall. Two weeks later, another wave of bombers attempted a second attack, but failed due to technical problems with their explosive devices. What most shocked the British population was that these were suicide attacks by their own nationals.2 Dr David Pepper, Director of GCHQ, observed that ‘What happened in July [2005] was a demonstration that there were conspiracies going on about which we essentially knew nothing.’ The British government now had to rethink its surveillance strategy.3
GCHQ was now fighting on three fronts. It was taking a leading role in the so-called ‘War on Terror’ while supporting British forces in two major conflicts, in Iraq and Afghanistan. In the context of the prolonged insurgencies that had developed in both countries, the new idea was the ‘Total Information Battlespace’ that sought to connect strategic intelligence-gathering from satellites, and every other kind of source, with soldiers on the ground. The old idea that GCHQ would mostly support the ‘high-ups’ in Whitehall while the troops fended for themselves with their own tactical sigint collection or ‘Y units’ in ageing Land Rovers had been completely abandoned. The challenge was to bring together sigint from many sources, and indeed from many intelligence services, since both campaigns involved numerous national contingents from places as diverse as Spain, Denmark and Ukraine. Sigint connectivity with the Americans was still a difficult issue. British, Australian and Canadian special forces, which were never shy when it came to speaking their minds, asserted that this problem had cost allied lives in the early part of the campaign in Afghanistan. Almost no sigint was shared with the Iraqi or Afghan forces, despite the efforts to rebuild them.
One of the most effective allied sigint operations was run by a Dutch intelligence chief who belonged to NATO’s International Security Assistance Force (ISAF) around Kabul. General Joop van der Reijn spent the months before his departure to Afghanistan joining up the intelligence practices of the Dutch and German forces. Once he had arrived, one of his most valuable innovations was improvised local sigint. He took over a building and filled it with low-cost ‘off the shelf’ radio intercept equipment that was commercially available. Two dozen former Afghan policemen were selected to become translators and eventually sigint monitors. Any problems with the rudimentary equipment were more than offset by the ability of this unit to comprehend the innumerable dialects and opaque infer ences in the Taliban conversations they intercepted. This unit took a little time to set up ‘given the Afghan way of doing business’, but eventually it became an invaluable part of the Joint Security Coordination Centre, which was ISAF’s intelligence hub.4
On the ground, much of the British sigint work in Afghanistan and Iraq was conducted by Light Electronic Warfare Teams, or LEWTs, from Britain’s last remaining dedicated field sigint unit, 14 Signals Regiment.5 Although it deployed the same vehicle-mounted equipment that had been effective in Yugoslavia, namely the Odette sigint system and the Vampire direction-finding system, much of the important work was done on foot at dangerously close range. The main instrument of the sigint war in Afghanistan was ‘Scarus’, a portable interception kit operated by four-man teams which looked rather like a backpack radio with oversized aerials. This was introduced in the summer of 2003, and by 2004 a dozen units were operational. The range of Scarus is only ten miles, and this meant forward monitoring operations. Scarus was often supplemented by commercially purchased hand-held scanners of the sort used by aircraft enthusiasts to listen in to air traffic control towers. Again this was improvised sigint, since these could be used most effectively by Afghan guides accompanying infantry patrols.6
British sigint operators often worked with their NATO colleagues. In 2006 they teamed up with Danish electronic warfare operators in Afghanistan’s Helmand province under the Sigint/EW Operations Centre in Regional Command (South). Their work earned them praise as ‘battle-winners’, providing real-time force protection and situational awareness. Many of the senior army commanders were experiencing the enormous potential of sigint for the first time. Intense operational pressure led to experimentation. The most remarkable achievement was linking up sigint aircraft circling overhead to the four-man LEWTs embedded with Parachute Regiment companies, and also serving the platoon-sized long-range patrols.
However, the cost was high. LEWT troops lost four men in action in the period up to January 2008.7 Among those killed was Lance Corporal Jabron Hashmi, the first British Muslim soldier to be killed on active duty in Afghanistan. He died along with another sigint specialist, Corporal Peter Thorpe, when they were hit by a rocket-propelled grenade during a firefight near Sangin. Born in Pakistan, less than an hour from the border with Afghanistan, Hashmi had come to Britain with his family when he was aged twelve. He had joined the Army and undergone specialist training at the Defence Intelligence and Security Centre at Chicksands before being posted to the Army’s main sigint unit. The local Parachute Regiment commanders spoke of the vital protection his s
pecialist work had provided to their patrols.8 GCHQ also sent significant numbers of people out to Iraq and Afghanistan: managers were surprised by the willingness of staff from Cheltenham to volunteer for these very arduous postings.9
The Royal Marines were prominent in the Afghanistan campaign, and had their own dedicated tactical sigint unit, in the form of Y Squadron from 3 Commando Brigade. Again the emphasis was on a front-line approach, with mobile Radio Reconnaissance Teams accompanying patrols. In May 2007 the Royal Marines located and eventually killed the Taliban’s most important chief, Mullah Dadullah, in an operation that was driven by sigint. He had been tracked across Helmand province by intercepting the calls he made to his brother on a satellite phone.10 By January 2009, commanders boasted of six thousand significant intercepts and in the region of seven hundred enemy call signs identified. The most significant contribution was advance warnings of enemy ambushes in Helmand which saved many lives. Once patrols understood what Y Squadron’s capabilities were, they wanted to take a Radio Reconnaissance Team with them every time they went out on patrol. Fusing sigint from tactical, operational and even strategic platforms was the new style of intelligence operations. The main vehicle for this was the recently formed in-theatre Sigint Electronic Warfare Operation Centre. Experiences in Afghanistan after 2002 led to a new overall sigint approach for NATO, tried out first in Operation ‘Trail Hammer’ in 2006. The objective was to share sigint seamlessly between GCHQ, the Nimrods of 51 Squadron, ships with dedicated sigint collection suites like HMS Cornwall, and forward sigint collectors. A high priority is extending this sort of support to units like the SAS.11
Circling overhead in Afghanistan were Nimrod R1 reconnaissance aircraft. Although the R1 sigint variants of these planes operate regularly over Afghanistan, the majority of Nimrods used there were maritime reconnaissance aircraft, originally designed to search for submarines. Given that Afghanistan is without benefit of a coastline, or even large stretches of water, this may seem peculiar. Their presence was required because until 2010, Britain lacked effective unmanned reconnaissance aircraft, or ‘drones’, of the sort operated by the Americans, since the type of drone initially purchased by the British, called ‘Phoenix’, was much inferior, and would not fly in hot weather or at the altitudes required for operations in Afghanistan.12 Indeed, when deployed in Iraq in 2003, one drone had been lost for every six missions flown. As a result, the ageing fleet of Nimrod maritime reconnaissance aircraft, six of which had some real-time video surveillance capability, was stretched to the limit serving as temporary drones and sigint collectors. One of these planes suffered a calamitous explosion in September 2006, with the loss of sixteen lives. In-flight refuelling of Nimrods was halted for a period, and then they were all grounded, leaving troops dangerously exposed.13
Belatedly, the British purchased larger and more effective foreign-made drones. The most important model was the Watchkeeper 450, which had an impressive ability to listen in to Taliban communications. This was an Israeli-made device that had been used extensively to conduct missile attacks over Lebanon in 2006, and later over Gaza. For the British, the Watchkeeper provided much-needed video coverage and also greatly improved sigint, since at altitude, interception range is much better. Currently, the elderly Odette intercept system is being replaced with ‘Soothsayer’, which offers greater inter-operability with the Americans and a much higher level of automation. Significant investment in front-line sigint equipment to serve the troops is, at last, being made.
While circling over Afghanistan, the RAF’s Nimrod R1s noticed something rather interesting. From 2005, it was increasingly common for these surveillance planes monitoring Taliban radio communications to come across militants speaking with Bradford or Birmingham accents. A case in point was Rashid Rauf, a Birmingham man sought by British police in connection with an August 2006 plot to bomb transatlantic airliners. He is believed to have escaped from a Pakistan prison in 2007, and was killed just inside Pakistan in a CIA missile attack on a militant stronghold in 2008. Intelligence officers observed privately: ‘He’s not the only British Muslim to die out here.’ Predictably, this raised fears that experienced fighters might have returned home from Afghanistan to plot attacks in Britain, and it was not long before RAF aircraft with similar listening equipment began circling over British cities searching for returned Afghan fighters. Their brief was to seek out suspects using ‘voice prints’ of fighters with British accents that had been collected by the Nimrods from Taliban battlefield communications.14
The surveillance effort between the war in Afghanistan and the ‘war’ at home was now seamlessly connected. The RAF had purchased three Britten-Norman Islander aircraft which were equipped with sigint suites and used as airborne listening stations. Based at RAF Northolt in west London, they are used for covert surveillance by MI5, which provides the monitoring staff in the aircraft. In early 2007 they were used to support the West Midlands Police when tracking suspects connected to a plot to kidnap and behead a British Muslim soldier. These aircraft can fly at fairly high altitude, and have been seen loitering over the East End of London for long periods. They are also known to operate out of Leeds-Bradford Airport. Many have presumed that they were deployed as part of a post-9/11 surveillance effort, but newspapers have published photographs of an MI5 officer standing next to one of them as long ago as 1999.15 Although run jointly by the RAF and MI5, like other ‘Home Office’ interception projects, this is partly a GCHQ activity, using technicians on temporary loan to the Home Office. It recalls Peter Wright’s ‘Airborne Rafter’ programme in the late 1950s, again run jointly by MI5 and GCHQ, which searched for KGB spies by tracking their radios.16 The residents in Muslim areas of Birmingham, Bradford, Leeds and London are not necessarily aware of the sigint flights, but they are highly sensitised to surveillance. ‘You are always looking over your shoulder here,’ remarked one local – even council workmen digging a hole in the road ‘means MI5, GCHQ, and the installation of monitoring devices’.17
These British monitoring aircraft use a technology that was developed during the drug wars in Colombia in the early 1990s, when the sigint agencies were involved in a technical war against well-resourced cocaine cartels. Once Pablo Escobar, the head of the Medellin drugs cartel, had been killed in December 1993, other drug lords became more anxious about the vulnerability of their cell phones. Their response was to frequently change their phones to confuse the eavesdroppers. Remarkably, they could even make use of scanners to steal the identities of the phones of passers-by, which were then cloned. The answer was a new aerial technology which depended on small spy planes. Cessnas and other light aircraft were fitted with directional antennae and wide-band recorders to take all the traffic from across the major bands on the entire spectrum used by cell phones. A single aircraft could take much of the mobile phone traffic of an entire city by circling at a high altitude and sucking up the microwave signals that connect the cell sites to the phone networks. The plane then pushed the product to a ground station where a massive computer examined the audio content of each conversation using voice-recognition software. This technique has not been effective everywhere, since some regions do not have much in the way of cell phone systems. However, in the skies over Britain this is now an important element in the armoury of the intelligence agencies.18
Comparing voices recorded over Britain with voices recorded over Afghanistan underlined the fact that the boundary between foreign wars and domestic terrorism was now perilously thin. In 2006 GCHQ launched two new programmes subsumed under the broader ‘SIGMod initiative’, or sigint modernisation programme. The exact nature of this remains classified, but it clearly focuses on the way in which GCHQ collects, analyses, presents and disseminates intelligence about terrorism. The anxiety that Britain’s oversight body, the Intelligence and Security Committee, has expressed about the need ‘to ensure that new interception techniques are regulated by a proper legal framework’ is another indication that the boundary between domestic and fore
ign communications is rapidly collapsing.19 SIGMod has resulted in a twentyfold increase in GCHQ’s ability to access, process and store material from internet-based communications, and ‘vast sums of money’ were spent on these two projects. Meanwhile, GCHQ decided to deploy yet more analysts to work alongside MI5 and other partners engaged in counter-terrorism.20
What was GCHQ’s own reading of its growing place in what had proved to be an American-led ‘War on Terror’? Detecting the broader political sensibilities of an entire secret service is not easy. However, by early 2006 there was a steady rise in staff asking difficult questions about legality, set against the background of high-profile issues such as CIA secret prisons in Europe that were spilling onto the front pages of Britain’s newspapers. Inquiries by the Council of Europe and the European Parliament into this matter prompted the Intelligence and Security Committee to look at British involvement in what were often called ‘extraordinary renditions’, in other words the secret apprehension and transfer of terrorist suspects without warrant. The heads of MI5, SIS and GCHQ were all asked for their views. When questioned, Sir David Pepper, GCHQ’s Director, seemed especially anxious to distance his organisation from the affair. On 29 November 2006 he explained that GCHQ had ‘never knowingly provided support to a US rendition operation and we would not authorise the use of intelligence for that purpose…and we have never been asked to do so’. He added that GCHQ’s only knowledge of the rendition programme ‘has essentially flowed from what SIS have learnt and told the other Agencies’. On rendition, he insisted, GCHQ had no independent source of information, and had merely ‘followed SIS’s growing understanding of what the U.S. was doing’. Pepper then outlined the standard guidance given to GCHQ’s analysts for situations in which they fear that unlawful behaviour might result from supplying intelligence to a foreign partner. This included possible upward referral, ‘ultimately to Ministerial level’. One gets a strong sense that not all aspects of the ‘War on Terror’ have sat comfortably with GCHQ’s management board.21
GCHQ Page 57
