GCHQ

by Richard Aldrich

  It was not until 8 December 1988 that Margaret Thatcher quietly announced that, after considering a new report by the Medical Research Council, and a further report by the Security Commission, the so-called ‘second phase of trials would not now go ahead’.72 This was a quiet admission of defeat. Everyone knew that the polygraph was unscientific and inaccurate. Its main value was that it deterred those with a guilty conscience. NSA had observed that the greatest benefits it provided were the things people confessed to before or immediately after taking a polygraph test because of the state of anxiety it created. Infamously, President Nixon once discussed the polygraph with his inner circle and offered remarks in much the same vain: ‘I don’t know anything about polygraphs, and I don’t know how accurate they are, but I know they scare the hell out of people.’73

  Was the government correct to seek to remove union rights at Cheltenham? On balance, the answer is probably yes. It is not widely understood that union issues had been a persistent problem over several decades, and the presence of the unions at GCHQ was anomalous compared to MI5 and SIS. In the late 1970s, as the highly secret J-Ops section of GCHQ became a processing component for NSA’s massive satellite ‘hoovering’ operations, the possibility of disruption in sensitive areas, such as round-the-clock computer processing, caused a collective neuralgic twinge on the part of intelligence chiefs. By that time GCHQ was not only being affected by local grading disputes, but also by national disputes in which the agency was seen as a useful ‘sore spot’ to annoy government. Something had to be done.

  However, Thatcher’s approach was incompetent. It put GCHQ in the spotlight, a place it was not supposed to be, and also damaged morale. Summarily removing trade union rights from those who already had them raised profound issues of principle. Michael Herman, the former head of J Division, offers a characteristically acute observation, noting that the sight of senior civil servants ‘marching in Cheltenham with the National Union of Mineworkers in the annual rally of protest sums up the effects’.74 There were several better options available. The First Division Association wisely recommended a continuity of service agreement. Alternatively, if government had been determined to move to a staff association, it could have taken a decision simply to allow no new union members. This, together with an offer of £1,000 to voluntarily relinquish rights to union membership, would have driven numbers down quickly. Hounding the few remaining unionists was unwise and vindictive.

  For GCHQ, the worst effect was upon staff commitment. The nature of its work made it peculiarly dependent on corporate spirit and collegiality. Nancy Duffton, a GCHQ worker, told the House of Commons, ‘Morale is very low since the union ban…The important thing, as far as the work is concerned, is that the sort of dedication that people used to have to the job is beginning to dwindle.’75 GCHQ workers told the Economist that morale was at ‘an all-time low’, and because of this the agency was working at only 80 per cent of its normal volume. In some of the outstations like Hong Kong, many of the Radio Operators just put their feet up and drew their pay. Most were agreed that the main fault lay with Margaret Thatcher. Whether right or wrong in principle, the government handled GCHQ with astonishing heavy-handedness and inflexibility.76 The arguments would drag on until 1997, when the trade union ban was reversed by the incoming Labour government of Tony Blair. Meanwhile, the secret activities of GCHQ received enormous publicity.77 Geoffrey Howe, in a remarkably honest assessment, conceded that: ‘Almost every aspect of the work and location of GCHQ was rehearsed again and again in the press. Our most secret service had become almost the most public.’78

  22

  NSA and the Zircon Project

  Thatcher said ‘we will strain every sinew’ to have Zircon.

  Peter Marychurch, Director of GCHQ, 1 May 19871

  On 24 January 1985 the American space shuttle Discovery was launched from Cape Canaveral. So far there had been fifteen shuttle launches, but this particular flight was unique. It was the first shuttle to be deployed on an intelligence mission, and it carried a highly secret new sigint satellite code-named ‘Orion’. Weighing almost six thousand pounds, the satellite was guided from the cargo bay by two military astronauts before being powered into its final orbit by rocket boosters. Once there, it unfurled two massive parabolic antennae that looked like huge umbrellas, each stretching out more than a hundred yards. One of these collected signals, including low-power radio transmissions that no sigint satellite had hitherto been able to hear. The other beamed the ‘take’ back to earth. The United States had been operating sigint satellites for more than twenty years, but this was by far the most powerful and impressive example, confirming the status of the USA as the world’s premier intelligence collector.2

  Sigint satellites, rather like fibre-optic cables and then personal computers, were part of an unstoppable world communications revolution.3 These technical breakthroughs had profound implications for sigint, and on both sides of the Atlantic intelligence chiefs recognised that this was changing the UKUSA alliance relationship. As a result of the growth of satellite collection, GCHQ was working more closely with NSA than ever before, notably by processing about 15 per cent of the ‘overhead’ material in its highly secret J-Ops section. Yet, paradoxically, GCHQ was also being left behind, and the underlying feeling was one of growing ‘unspecialness’. Indeed, with the possibility of machine translation beckoning, there was even a danger that the Americans might eventually view GCHQ as expendable. Therefore, in the 1980s both GCHQ and NSA were reconsidering their intelligence alliances, not only with each other, but also with the long-established ‘second party’ members of the original UKUSA agreements such as New Zealand, and even with the ‘third party’ sigint services in Western Europe, such as Sweden’s FRA and Germany’s BND.

  Any venture into space meant tough political as well as technical choices. In the mid-1960s Harold Wilson’s Cabinet had opted for the British-made Skynet military communication satellite, instead of GCHQ’s preferred option of an American-made model. Whether to buy cheap and reliable from the Americans, or to invest in expensive British national capacity (and jobs), or indeed even to join with the Europeans, was a perennial issue. The same dilemma of dependence or independence presented itself in other high-tech areas such as nuclear weapons and computing, as had recently been illustrated most starkly by a further British venture into space code-named ‘Chevaline’. This was a programme to upgrade Britain’s Polaris nuclear missiles by improving the final stage of the rocket – in effect a small space-ship – that carried the warhead on the last part of its journey towards its target. Chevaline was the direct precursor of Britain’s remarkable adventure with a sigint satellite called ‘Zircon’ in the 1980s.

  Britain’s main nuclear deterrent consisted of American-made Polaris missiles carried on submarines. No sooner had Britain taken delivery of Polaris than it started to look outdated. The Russians and Americans had begun to experiment with elaborate anti-ballistic-missile systems to defend their cities, threatening a new spiral in the arms race. Unless Britain’s Polaris missiles were either replaced or upgraded, they would most likely be defeated by Moscow’s anti-missile defences, rendering Britain’s independent nuclear deterrent implausible. The ability to hit Moscow was considered to be the key criterion for an effective British deterrent.4 Throughout the early 1970s British defence chiefs pressed for the purchase of a new American missile called ‘Poseidon’, with multiple warheads powerful enough to overwhelm Soviet defences. One of the reasons they wanted to buy American was because they had always recognised that ‘Without US intelligence support any nuclear deterrent system would lose credibility in a few years.’5 In other words, access to American sigint was vital for targeting British nuclear weapons.

  However, Britain’s political leaders had recognised that there would be a diplomatic furore if Britain bought Poseidon. The country had just entered the Common Market, and the French would certainly not appreciate such a public reaffirmation of the ‘special relationship’ at t
his crucial moment. Accordingly, in late 1973 the Edward Heath government rejected the Poseidon option and chose a British programme to update and improve the ageing Polaris missiles, code-named ‘Chevaline’. At the time, officials believed that this code name was French for a nimble mountain goat. In fact, when the Chevaline saga unravelled some years later in the House of Commons, it turned out that its true meaning was either someone who sells rotten horse flesh, or an ugly woman. There was considerable humour at the government’s expense.6

  Britain’s defence scientists were delighted by the Chevaline decision, which effectively gave them their own mini-space programme and a substantial secret budget for advanced science. Some defence intelligence experts were less impressed. Louis Le Bailly, Director General of Intelligence at the Ministry of Defence, was working closely with the CIA, and enjoyed the best insights into the Russian anti-ballistic-missile systems which Chevaline was supposed to defeat. He was not optimistic about its chances: ‘I personally laboured with my advisers for the best part of six months to transpose this deeply technical data into layman’s language so that my colleagues on the JIC could understand its impact, which I believed to be profound.’ However, at the last moment senior scientists persuaded the Cabinet Secretary to have intelligence reports criticising the Chevaline decision withdrawn from circulation. The chairman of the Joint Intelligence Committee was reportedly ‘almost prostrate with hysteria’.7 In protest, on 17 July 1974 Le Bailly handed in his resignation to the Cabinet Office Intelligence Coordinator. Chevaline was quite simply the hottest defence issue of the decade.8

  It was also very secret. In March 1974, when a Labour government took over from Edward Heath, Harold Wilson was anxious to hide the project from his own backbenchers, who were not at all sympathetic to nuclear weapons. Indeed, Wilson even hid it from some of his Cabinet Ministers. Therefore, despite the huge sums involved, the existence of Chevaline was not revealed to the House of Commons Public Accounts Committee, or indeed reported to Parliament in any way.

  Meanwhile, the super-secret programme was not going well. For years Chevaline, which was effectively a mini-space capsule that provided a new front end for Britain’s Polaris missiles, was test fired but failed to work properly. The cost escalated from an initial £175 million to £600 million, and then £800 million. In 1982, when the highly secret programme was finally exposed, there was an almighty row, not least because both front benches had been involved in an elaborate deception of Parliament. Officials had to confess that they had conspired to spend close to a billion pounds in complete secrecy on a failed project. This was a major public scandal, since Parliament was supposed to control major expenditure. Shortly afterward, a special ‘Never Again’ agreement was drawn up, with officials promising faithfully not to hide large defence projects from Parliament.9

  The ‘Never Again’ agreement did not last long. The cause of its demise was GCHQ’s desire to venture into space. By late 1983, Britain’s alliance relationships were being re-examined in the light of the recent Falklands War. Once Alexander Haig’s effort to find a peaceful compromise had ended in late April 1982, American military assistance had flowed towards the British in vast quantities. The rapid provision of the latest American Sidewinder air-to-air missiles for the Royal Navy’s Harrier jets alone had done much to decide the war in Britain’s favour. Casper Weinberger, the American Secretary of Defense, had gone so far as to say that if Britain lost an aircraft carrier, he would happily provide a replacement. The same applied broadly to intelligence assistance, and during the war some of the corridors in the CIA headquarters at Langley in Virginia were actually decked out in Union Jack bunting. Yet in Cheltenham, GCHQ’s evaluation of the ‘special relationship’ was somewhat mixed. Imagery satellites were not especially helpful because of the almost complete cloud cover over the Falklands during the South Atlantic winter. For sigint coverage, GCHQ had pleaded for the use of an NSA listening satellite that was then focused on Central America, supporting the Contra struggle against El Salvador and Nicaragua, which was of high importance to the Reagan administration. In order to cover the Falklands the satellite had to be moved, and NSA was willing to do this for only a few hours a day. Persuading it to do even this much had taken a good deal of argument on the part of GCHQ’s Director, Brian Tovey.10

  GCHQ now had to face some unpalatable facts. It had been left behind in space, which was clearly the future of sigint. Given that sigint provided more than 80 per cent of Britain’s intelligence, and took a similar proportion of the budget, this was a serious matter, conjuring up the possibility that during some future crisis the United States might be unwilling or unable to divert its resources to help Britain, either because they differed on an issue politically, or because NSA was preoccupied elsewhere. All the senior staff at Cheltenham recalled Kissinger’s sinister attempts to ‘cut off’ the intelligence flow to Britain in August 1973, and even now a similar spat was looming between the United States and New Zealand. Thus, in his last year as GCHQ Director the energetic Brian Tovey proposed that Britain should launch its own sigint spy satellite, code-named ‘Zircon’. He reasoned that GCHQ was judged to have performed well during the Falklands War, so this was the right moment to press Margaret Thatcher for a substantial reward. Zircon was bound to be very expensive. Like the Nimrod R1 in the 1970s, it could not be accommodated within the routine ‘Cost of Sigint’ budget. Some have estimated that the satellite and its associated ground installations might have cost £500 million over five years, and thereafter it would need replacing.11

  Many believe that Zircon had originally been conceived by Brian Tovey ‘to keep the special relationship sweet’ and to take his organisation into space.12 However, others have read it as a deep questioning of the ‘special relationship’. As Bill Odom, Director of NSA after 1985, put it, ‘You do not invest all that money in a satellite system if you believe that the Americans will continue to give it to you for free. That would be plain stupid.’13 Some believed that GCHQ’s satellite venture was part of a radical new strategy. Tovey was sympathetically disposed towards some of the European ‘third parties’, such as Germany and Holland, and had provided them with more material than was generally realised. In particular, he enjoyed a good relationship with the charismatic Alexandre de Marenches, head of the French secret intelligence service, SDECE.14 It is not inconceivable that Zircon was about GCHQ becoming the biggest fish in a European pool. A sigint satellite like Zircon, working in cooperation with the French satellite imagery programme, would have made Europe a serious intelligence collector. At the very least, Zircon guaranteed Britain an alternative option at a time when the European ‘third parties’ were becoming more powerful and influential.15 Indeed, the Europeans had recently set up their own mini-UKUSA alliance called ‘The Ring of Five’, consisting of the sigint agencies of Germany, the Netherlands, France, Belgium and Denmark – although this did not prevent them from intercepting and reading each other’s communications traffic.16

  The idea of GCHQ teaming up with the Europeans was not a bad one. European sigint services such as those of the Norwegians, the Germans and the Dutch were highly professional, and had made many important contributions to Western intelligence down the years, not least during the Falklands War. They often displayed brilliant lateral thinking. When Soviet naval ships entered Dutch harbours in the 1980s, their hosts would often complain that the Soviet radar was interfering with local television broadcasts and insist on a snap technical inspection. The crafty Dutch used this opportunity to plant a small tracking device high up on the Soviet ships that was no more than the size of a brick. This tracking device was so successful that it became a standard technique across the Western secret services. Typically, a small group of British SBS personnel worked with SIS and GCHQ on similar tasks in the 1990s.17

  GCHQ’s relationship with its American partner was about people as well as projects. General Bill Odom arrived as the new Director of NSA in March 1985. Odom was a tough-talking Army officer with an extremely
abrupt manner. He saw himself as a new broom, complaining that his predecessor ‘would not favour radical change’ and that the staff at NSA were ‘too laid-back’.18 He also looked afresh at the Anglo–American sigint relationship, and was deeply unimpressed, observing that ‘The name of the British game is to show up with one card and expect to call all the shots.’19 Ingenious old-fashioned British cryptanalysis was being overtaken by the raw power of America’s Cray super-computers, and this had been underlined by some remarkable NSA breakthroughs with Soviet high-grade diplomatic traffic in the late 1970s.20 Odom noted that, ‘What the British brought in World War II, they do not bring any more…Today, this business requires huge investment, and Britain doesn’t have that.’ Britain’s decision to buy Zircon signalled GCHQ’s renewed commitment to spend big money on sigint.21

  Bill Odom’s first year as Director at NSA was a traumatic one. Washington soon dubbed 1985 ‘the Year of the Spy’, since it brought the exposure of Ronald Pelton, a damaging mole who worked in ‘A5’, NSA’s sensitive Soviet section. With a photographic memory, Pelton proved even more disastrous to Western code-breaking than Geoffrey Prime. At almost the same time the Walker family, a whole group of spies working within US naval intelligence and communications, was uncovered. Edward Lee Howard, a CIA officer, was also revealed to be working for the Russians. To cap it all, in November 1985 it was discovered that Jonathan Pollard, a Mossad spy inside the Pentagon, had been handing over very sensitive material to the Israelis. The Americans had not yet uncovered the two best covert sources employed by the Soviets, namely a CIA officer called Aldrich Ames and an FBI officer, Robert Hanssen.22 These frightening cases of KGB espionage had a direct impact on the British, since they made Bill Odom all the keener to see the polygraph deployed by GCHQ at Cheltenham.23