Katharine Gun’s decision to expose the NSA message was taken on the spur of the moment. On Friday, 31 January 2003 she had gone to work at GCHQ as usual. At about 10 o’clock she opened her emails. ‘I could not believe what was on the screen,’ she said. ‘My thoughts were racing, really bizarre thoughts for me. I had never intended to do anything like that…’ In other words, she had not been looking for material to leak, but now she felt she was privy to the ‘most secret workings of top government’. Moreover, it struck her that this document – if leaked – might well be used to impede military action against Iraq. Gun’s critics have since denounced her for naïvety, asserting that someone working for GCHQ should have understood that sigint agencies monitor everyone, including friends and neutrals, even the United Nations. Her response to this was that the memo exposed a lie. London and Washington claimed to be working for a diplomatic solution, but in fact they seemed to be trying to avoid one.40 They also seemed to be trying to manipulate the key vote in the Security Council by ‘bullying’ the smaller members. She recalls being horrified and angry.41
The Monday morning after the story broke, GCHQ began an immediate leak inquiry, interviewing over a hundred staff who had seen the email. Gun had not expected the Observer to reproduce the entire text on the front page, and had been ‘absolutely terrified’ when she saw it. Although she denied her action during her security interview, a few days later her nerve crumbled and she confessed to her line manager. She was taken to GCHQ’s Security Division, then interviewed by Special Branch from London. She never returned to GCHQ.42 The case, which ran for over a year, was headline news. To her supporters she was the ‘the spy who tried to stop a war’. Others were less complimentary. David Blunkett, the Home Secretary, insisted that the NSA memo had been ‘doctored’ before publication, and believed that Gun was motivated by the fact that her husband, a Kurdish asylum-seeker, was being removed from the country.43
No less significant than the leaked email from NSA was another message that had been sent to all staff in GCHQ the previous week. The issue of possible war against Iraq was causing growing anxiety among staff at Cheltenham, and a senior official had tried to address their concerns, assuring them that they would not be asked to participate in anything unlawful, and that British troops would not go into action unless the Attorney General, Lord Goldsmith, had advised the Prime Minister that it was legal. This was to prove important nine months later, since Katharine Gun’s acquittal of charges under the Official Secrets Act turned precisely upon the Attorney General’s legal opinion on military action. Because of the GCHQ email underlining the importance of lawfulness, Gun’s defence team asked to see the full text of Goldsmith’s opinion on the legality of the Iraq War. At this point the government’s lawyers crumbled.
Despite the fact that Gun had admitted her actions, the Crown Prosecution Service dropped the case ‘for lack of evidence’ within twenty-four hours. The government did not wish to reveal Goldsmith’s full legal opinion under any circumstances, since it was equivocal. This in turn underlined the fact that journalists had been barking up the wrong tree with their obsession over the Iraq dossiers, their testy arguments with Alastair Campbell and the supposed fact that Saddam would be capable of deploying his WMD at forty-five minutes’ notice. The Katharine Gun case showed, rather belatedly, that some of the bigger issues relating to the approach of war had been missed by the press.44
GCHQ monitoring of the United Nations remained stubbornly in the headlines. In February 2004, Clare Short, who had resigned her Cabinet post as Minister for Overseas Development, offered her own testimony about listening in on the UN, declaring on prime-time television that she had routinely seen sigint on Secretary General Kofi Annan’s conversations that had taken place in his private office at the UN headquarters in New York during the period before the war. Just as with Mo Mowlam in Northern Ireland in the 1990s, the whole matter of interception had a rather surreal quality. Short said: ‘I have seen transcripts of Kofi Annan’s conversations. In fact, I have had conversations with Kofi in the run-up to war thinking, “Oh dear, there will be a transcript of this, and people will see what he and I are saying.” ’45 This was embarrassing, since it not only showed that GCHQ had been intercepting Annan’s phone calls and emails, but also that a clandestine listening device had been surreptitiously planted in his office.46
As war drew closer, Clare Short took an unusual step. ‘I had decided that I ought to inform Kofi that transcripts of his conversations as well as draft papers were circulated by British intelligence.’ This monitoring had seemed harmless when Britain was working closely with the Secretary General, but now she considered that it had become ‘insidious’. When the Katharine Gun case erupted, Short spoke on BBC Radio 4’s Today programme and revealed that ‘We were also spying on Kofi Annan’s office.’ Following this interview, she recalls, ‘All hell broke loose,’ and she received a warning from Tony Blair.47 However, she was not deterred. On 29 February 2004 she was a guest on the Jonathan Dimbleby programme on ITV. Here she displayed a letter from the Cabinet Secretary, Andrew Turnbull, rebuking her for discussing signals intelligence in public and threatening her with unspecified ‘further action’ if she did not stop talking about GCHQ.48
GCHQ’s most important potential military contribution to the war against Iraq was in its first few hours. Reportedly, late on 19 March 2003, Cheltenham discovered that Saddam Hussein was meeting senior commanders at a house called Dora Farms in the prosperous Mansour district on the southern outskirts of Baghdad. A number of special forces personnel from Britain and the United States were already operating inside Baghdad, as were two members of Germany’s BND. They were asked to confirm the information. At about 6 o’clock on the morning of 20 March, Baghdad time, the opening of hostilities saw two American F-117 bombers hammer the area, dropping four two-thousand-pound bunker-busting bombs, leaving a crater sixty feet deep.49 Two weeks later, the British asserted that they had intelligence from GCHQ that Saddam left the area minutes before the bombing, but had been injured. However, the CIA remained ‘cautiously optimistic they got him’. Reporters later discovered that the bombs had missed their target, and had instead flattened three neighbouring buildings. Reportedly, the strike was triggered by the Iraqi leadership’s use for communications among themselves of an old system made by the British company Racal which had been monitored.50
Accompanying Britain’s land forces were two mobile sigint units, in the form of the Army’s 14 Signals Regiment and the Royal Marines’ Y Troop. Most of the staff of Y Troop flew out to Kuwait in the last week of January 2003. Usefully, some of them had recently been on an eight-week Arabic course. They were located at a Kuwaiti training camp, allowing some of their detachments to be deployed to Mutlar Ridge, overlooking the Iraqi frontier, which allowed them to test equipment and to liaise with the US Marines’ 1st Radio Battalion, which was also a field sigint unit. After about a month, Y Troop spread out along the border in small detachments to listen to the enemy and await instructions to advance. Their main equipment was three Odette sigint systems, mounted on Land Rovers with their distinctive twenty-foot aerials. However, there were technical problems, so most of Y Troop was sent forward with smaller man packs as radio reconnaissance teams, working alongside 40 Commando. This was front-line work, and they were soon amongst the tanks of the Queen’s Dragoon Guards, listening in on enemy positions at very close range as British armour advanced towards the southern town of Basra.51
Eventually the technical problems with the Odette aerials were fixed by 14 Signals Regiment, the Army’s main electronic warfare unit, allowing Y Troop to go back to longer-range intercept work. At this point a small Iraqi patrol found them, and managed to launch some RPG rockets at their encampment before being seen off with a barrage of automatic fire. There were no casualties, but it was a sharp reminder of the hazards of forward interception. Thereafter, the number of targets they were tracking became progressively smaller as Iraqi soldiers deserted their positions. A
notable discovery was the location of their counterparts, the Iraqi 124 Electronic Warfare Regiment. It was noted with satisfaction that this became ‘124 Crater Regiment’ after the application of well-directed artillery fire.52
As the British First Armoured Division advanced towards Basra, sigint provided timely warnings of ambushes that had been set up by Saddam’s Fedayeen guerrillas. It also obtained good information on the movements of key Iraqi leaders inside Basra itself.53 Sigint performed well in Iraq because many of the old problems of intelligence support to the front-line soldier had been solved. For the first time in perhaps fifty years, sigint flowed freely from national assets down to operational units, a stunning breakthrough when set against the history of patchy GCHQ support for previous major campaigns. Instead, the problem was sharing between allies. The Americans were well-equipped with pilotless drones that collected both imagery and sigint, of which the British felt the lack. As the initial invasion turned into a gruelling insurgency there was also a shortage of personnel who were experienced in running human agents. Some British sigint was being fed into American databases which then proved to be ‘for American eyes only’, causing frustration at the front line. There was no sharing of sigint with minor allies. In fairness to British and American intelligence agencies, while they had performed poorly on the issue of Iraqi WMD, they had accurately predicted the difficulties of a hostile occupation and the prolonged insurgency.54
As the Iraq War commenced, GCHQ’s Director, Sir Francis Richards, was preparing to depart to become Governor of Gibraltar. After a succession of three ‘outsiders’ as Director, GCHQ was being handed back to one of its own, the fifty-five-year-old Dr David Pepper. Joining GCHQ in 1972 after completing a D.Phil. in physics at Oxford University, he had risen to be Director of Administration by 1995, and served as ‘aide de camp’ during the Roger Hurn review.55 Pepper not only had a new role, he also had a brand-new office, located behind the gleaming glass doors of ‘the Doughnut’, GCHQ’s impressive new headquarters at Benhall. The first wave of staff was preparing to move in during August 2003, and senior officials assured reporters that even the builders had gone through security checks: ‘We didn’t want a repeat of the American Embassy in Moscow, which was riddled with bugs.’ The building was secure, but unfortunately it was not big enough. Planned during the 1990s contraction of intelligence, it was now too small to hold all of the GCHQ staff, which with the surge against terrorism was now approaching 5,300. About five hundred of these could not be permanently accommodated, so some buildings were retained on the old Oakley site, and ‘hot desking’ was the order of the day.56
‘The Doughnut’ had proved to be surprisingly troublesome. In late 1999 there had been dire warnings about a massive overspend on its computer equipment, and as a result the Cabinet Secretary, Richard Wilson, had asked Lieutenant General Sir Edmund Burton to review every aspect of GCHQ’s business management.57 Burton had some good ideas, but he could not fix the IT overspend which had triggered the review. When GCHQ planned its new building in the late 1990s, its technicians were still mentally living in the late 1980s. As David Pepper later explained, at that time its computers tended to be standalone systems, often quite large, but not interconnected. This reflected the fact that GCHQ’s main adversaries were also ‘large, monolithic, essentially static targets’ like Russia. By the mid-1990s the organisation was faced with much more volatile targets, requiring flexible approaches, and this accelerated GCHQ’s move towards computer networking. Yet the planners still envisaged a simple box-move into the new building, which meant switching everything off. Nobody stood back and said, ‘Just a minute. It’s not going to work like that, because of the degree of networking.’58
GCHQ’s original guess at the cost of moving its computers had been £41 million over two years. It was now a breathtaking £450 million.59 Parliament was incensed, and MPs were doubly angry because this kind of mistake had happened many times before. The computer costs of the new MI5 and SIS buildings had also rocketed. There were other precedents. The most obvious was a secret defence programme called ‘Project Pindar’, a top-level command bunker hidden under the southern end of the Ministry of Defence main building.60 This vast complex also included facilities for ‘Cobra’, the Cabinet emergency committee room.61 Although partly a civil project, the costs were buried within the Ministry of Defence budget.62 The agreed cost eventually quadrupled, and much of this had to do with the computers and communications equipment. There had been a similar experience with a disastrous Defence Intelligence Staff computer scheme called ‘Trawlerman’, which had to be abandoned.
In 1996, MI5’s computerised system code-named ‘Grant’ had also failed miserably, and was scrapped at a total cost of over £20 million.63 Indeed, there were so many precedents that MPs could not begin to understand why officials had not learned from this string of disasters.64
GCHQ’s mammoth overspend caused real friction within the intelligence community, since MI5 and SIS feared that their budgets would be squeezed to make up the cost.65 In the event, David Pepper was saved by circumstances, since the increased tempo of counter-terrorist activity made the case for uninterrupted operations unanswerable. Nevertheless, in the autumn of 2003 Parliament discovered that GCHQ had wasted further millions on new systems that had been bought to intercept terrorist communications. A costly prototype of a system that GCHQ hoped would increase its ability to listen in to terrorist traffic had only ‘partly delivered’.66 This was coded Whitehall language for ‘The computer system has failed and has been scrapped.’ MPs complained that ‘Security cannot be used as a smokescreen for incompetence.’67 Notwithstanding this, it has to be conceded that operational continuity at GCHQ had proved valuable in the recent past. On 24 January 2000 a catastrophic failure of NSA’s powerful computers halted the processing of American intelligence at Fort Meade in Maryland for more than three days. GCHQ took up some of the strain, and American customers were served directly from Cheltenham.68
By the summer of 2003, no one was taking much notice of GCHQ’s financial misdemeanours. Instead, the hot intelligence issue was the ‘missing’ Iraqi WMD, which had been the main justification for going to war. Four months after the invasion, no WMD had been found. The decision to invade Iraq had been politically divisive, and the patience of the British public had run out. There followed an unprecedented ‘season of inquiry’ into the secret services, with no fewer than four British investigations into intelligence aspects of Iraqi WMD between July 2003 and July 2004. The issues were looked at by the Parliamentary Select Committee on Foreign Affairs, and then by the Intelligence and Security Committee. A third inquiry, chaired by Lord Hutton, looked into the death of Dr David Kelly, a government scientist who had been a key defence intelligence expert, and who was closely cross-examined during the first inquiry. None of this assuaged the intense public anger, so finally, in early 2004, Lord Butler, who had been Cabinet Secretary under John Major, was called in to conduct a much wider investigation into British intelligence and WMD over the previous decade.
The controversy persuaded veteran Cheltenham staff to break their customary silence. In a letter to The Times in February 2004, Douglas Nicoll, the senior GCHQ officer who had previously reviewed the JIC’s performance on attack warning, denounced Tony Blair’s recent explanations before Parliament about how the JIC had interacted with other parts of the intelligence machine over Iraq as having ‘the highest degree of improbability’, and asserted that ‘politicos’ from No.10 had become involved in the intelligence process. Nicoll considered that this was ‘unprecedented’, and that there was much for Lord Butler to investigate.69
Everyone was asking the same question. Was the Iraqi WMD fiasco a product of intelligence failure by the agencies, or of deception by politicians and spin doctors? Inevitably, the answer is ‘both’. Having badly underestimated Iraqi WMD stocks prior to the First Gulf War in 1991, intelligence officers did not want to be caught out a second time, and so opted for ‘worst-case analysis’.
In other words, they over-corrected. Moreover, the allies cooperated so closely on WMD estimates that, far from challenging each other’s findings, they succumbed to a form of ‘groupthink’. Only the Dutch and Canadian intelligence communities expressed serious doubts. Butler revealed that a lot of British intelligence had come from a handful of human agents run by SIS who were not properly ‘validated’, and whose material was mediocre at best.70 The awkward evidence still continues to accumulate. Tyler Drumheller, the CIA Europe Division Chief, who worked closely with Britain’s SIS, has revealed that Naji Sabri, Iraq’s Foreign Minister, did a deal to reveal the country’s military secrets. Drumheller recounts that once policy-makers learned what Sabri had to say – that Iraq had no active WMD programme – ‘They stopped being interested in the intelligence.’71
There was also blatant political dishonesty. The British government had made three assertions in its WMD dossier. First, that there was plausible intelligence to suggest that the Iraqis might have hidden some old biological or chemical stocks from 1991. This was true, but these weapons were unlikely to be usable, and were at best of historical interest. Second, that there was evidence that Iraq continued to seek nuclear components on the world market, and nurtured future ambitions. This was also true, but no one thought for a moment that the country had got far with reconstituting its nuclear programme. Everything turned on the third claim, that Iraq was engaged in continued production of WMD. This assertion was made forcibly by the Prime Minister in his personal foreword to the Iraqi WMD dossier in September 2002. There was no credible evidence for this.72 Later, Blair assured the House of Commons that the intelligence was ‘extensive, detailed, authoritative’. This statement was also deeply misleading.73 Equally, Butler noticed that there was no change in the intelligence reports on Iraqi WMD during the period between 2002 and 2003, when the British government shifted dramatically from a policy of containing Iraq to one of confrontation.74 Alastair Campbell summed this up best in his diary, noting that the hardest question was: ‘Why now? What was it that we knew now that we didn’t before that made us believe that we had to do it now?’75
GCHQ Page 56
