In 1961, David Kahn, a budding author on the subject of cryptology, planned to publish a book on the highly secretive American Sigint (signals intelligence) agency, the National Security Agency (NSA), eventually entitled The Codebreakers. NSA leadership figured that its best hope to limit the perceived damage to national security was to pressure the publisher to scrap the whole project, comprising several hundred pages of material. When it became clear this was not going to happen, NSA scaled down its original request and focused on the deletion of three particularly sensitive paragraphs detailing its most valuable secret: NSA’s hand-in-glove collaboration with Britain’s own signals intelligence agency, the Government Communications Headquarters (GCHQ). Indeed, it appears that GCHQ leaders were as concerned as their American cousins over Kahn’s findings and dispatched their own top-ranking liaison officer to intercede with the publisher’s London office. Ultimately, the transatlantic pressure worked and Kahn was persuaded to drop the offending paragraphs.
The three paragraphs in question, later released in any case, would have informed the reader that NSA maintained close contact with GCHQ, including the exchange of liaison officers, and that these liaisons exchanged crypt-analysed material and techniques. Furthermore, the reader would learn that the two agencies geographically divided up the cryptanalysis workload so that NSA had responsibility for certain countries, GCHQ for others, and that the pair would exchange information to save duplication of effort, as well as exchange personnel to deepen the expertise and broaden the experience of their employees.
As this vignette illustrates, NSA and GCHQ believed their most jealously guarded secret, above all others, was the synergy of joint Anglo-American Sigint. Indeed, it is fair to characterise their Sigint alliance as the backbone of the special intelligence relationship. This synergy (or force multiplier) enables not only Sigint agencies, but all intelligence bureaucracies, to cover more languages, more material, offer more strategic indications and warning, and expend more money than a single country’s intelligence apparatus could do alone. This relationship also laid the basis for what would become the most effective liaison partnership that has ever existed between national intelligence bodies: the Five Eyes.
The nature of Sigint
Signals Intelligence may be the most prominent, encompassing, and important area of Anglo-American joint intelligence. That early cooperation began in the realm of Sigint stands to reason. Sigint is distinctive – even unique – in the US/UK context and worth exploring for several reasons. First, it is by far the most resource intensive method of intelligence collection in terms of infrastructure, cost, and labour. The development and maintenance costs of intercept stations on the ground or satellites in orbit is tremendous and the number of personnel required to process, translate, and analyse Sigint is correspondingly huge. It is largely for this reason that the UK and US first opted to cooperate closely in the Second World War and to continue this arrangement during the Cold War, even in the face of inevitable friction and occasional political disruption.
The importance of Sigint during the Second World War is well documented. American naval success at Midway and shared victory in the Battle of the Atlantic are but two classic examples. Its wartime reputation secured, Sigint remained the most treasured form of intelligence well into the Cold War as well. Part of this was owed to Sigint as an ‘INT’ distinct from other intelligence types. Human intelligence in the early Cold War did have notable successes, but these exceptions proved the rule that Humint operations during the early Cold War were ad-hoc, insecure, short-lived, and mostly involved debriefing defectors and travellers whose access to secret information was limited. It took decades before agents could be securely run in-place.
Imagery Intelligence (Imint), on the other hand, was effective in targeted missions, but dangerous, provocative and politically problematic before satellites replaced aircraft; an innovation which fundamentally changed Imint from the mid-Cold War onwards. Further, Imint was notoriously bad at discerning the plans and intentions of the Kremlin. Even open-source information was not only hard to come by, but also controlled by state media outlets and therefore unreliable. Thus, not only due to Sigint’s own inherent capabilities, but also due to a process of elimination, Sigint shouldered the intelligence burden during the early Cold War. This was clearly understood by all parties to the 1946 UKUSA agreement who noted, ‘The value of Communication Intelligence in war and peace cannot be over-estimated; conservation of the source is of special importance.’
If Sigint is of ‘special importance’, it is also the most consistent and formally integrated example of intelligence synergy and joint operation. The nature of Sigint cooperation requires a nearly all-or-nothing approach that is not true of Humint; in contrast to the CIA, for example, there is even a GCHQ liaison office on the executive floor of the NSA. Indeed, the default operating rule in Humint is unilateral unless specific operations are pursued jointly. In contrast, the Sigint approach is the opposite — share the load and divide the work. This philosophy has both guided and shaped the transatlantic Sigint alliance. This courtship process, and its conversion into a wider network of intimate collaboration, is worth exploring as a pre-history to its contemporary operations and future potential.
Origins of the Five Eyes partnership
When studying the genesis of the Five Eyes partnership, consideration of its wartime Sigint roots is required. The UK and US set up many bilateral agreements to decrypt, process, and share signals intelligence. Such agreements were expedient but fragmented, embracing various US and British entities such as the individual military services operating radio intercept stations. In 1943, an eventual separation was anticipated; thus, independence in Sigint expertise was considered the key to post-war intelligence success. Yet, ironically, a parting of the ways never occurred, and the oft-feuding American military branches sought their own Sigint independence from each other more than from the British. Indeed, it would be incorrect to conclude that the US military services branches sent liaison personnel to Bletchley Park solely because of a core belief in the value of cooperation with the British.
If in 1943 the Americans sought post-war Sigint independence from the British, by early 1945 this changed, and longer-term agreements were hammered out. Both sides realised that liaison was mutually beneficial, with a shared pool of intelligence supplied to different wartime theatres. The first substantive move was in October when a British delegation travelled to Washington to formalise a historic agreement, broad in scope and impact. As economic pressures pared back wartime budgets, it was clear that a post-war unified Sigint framework was needed to continue to harness the benefits of force-multiplication. Therefore, on 5 March 1946 the successor and heir to all wartime Sigint agreements, was signed into existence, known at the time as BRUSA. The parties agreed, ‘It was hoped that we could achieve [peacetime efficiency gains] by suitable division of labour and by avoiding duplication.’
The signatories agreed to a broad spectrum of Sigint cooperation; the most important aspects dealt with security, common operation, and personnel issues. The two sides agreed ‘to the exchange of the products of the following operations relating to foreign communications: (1) collection of traffic, (2) acquisition of communication documents and equipment, (3) traffic analysis, (4) cryptanalysis, (5) decryption and translation, and (6) acquisition of information regarding communication organizations, practices, procedures, and equipment.’ To affect this, an agreed system of security and codewords was established, joint communications channels were rapidly developed, and key installations became jointly manned. Further, liaison officers were posted to the other’s capitals. Both sides agreed, ‘Each party shall maintain, in the country of the other, at an appropriate location, a senior liaison official accredited to the other, who may be accompanied by the necessary staff. All Communication Intelligence liaison matters in each country shall be under the cognizance and control of the senior Communication Intelligence liaison official in that country.’ This agreement, often called the ‘Secret Treaty,’ laid the framework for cooperation that exists today.
If pooling of labour and common handling procedures were half of the synergy equation, division of global coverage was the remainder. Conceptualising the Sigint value of the post-war British Empire, it is well to remember its geographic expanse. GCHQ established a wide network of Sigint stations in Sri Lanka, Malaya, Hong Kong, Oman, Crete and Cyprus, and had a strong influence in the early Sigint operations in Canada. The English-speaking Commonwealth was included in the British Sigint order-of-battle during this time, especially their Australian counterparts, upon whose development British intelligence officers left a substantial mark. Key leadership positions in the nascent Australian Defence Signals Bureau were provided by GCHQ, which also maintained technical oversight of Australia’s Sigint operations from London. Mirroring bilateral Canadian-American wartime links, the Australians and the Americans also had an intelligence connection since early 1942 when General Douglas MacArthur arrived in Melbourne to assume command of the South West Pacific Area.
Australia was a friendly Sigint outpost and wartime partner for both the Americans and the British in Asia, although even here concerns over Soviet espionage interfered with the transition into peacetime cooperation. The post-war BRUSA agreement was thus first extended to include Canada, and eventually to Australia and New Zealand in the aftermath of the Korean War, taking the moniker of UKUSA along the way and forming what we now know today as the ‘Five Eyes.’ Other newly independent colonies in the British commonwealth were not as ready as Canberra or Wellington to develop independent Sigint links with the Americans.
Fortunately for Washington, the UK’s historical legacy in a number of strategic and convenient locations meant that British officials had the license to undertake local Sigint activity, and America could receive the intelligence product via their liaison agreements even if no American personnel or collection platforms were present. Calder Walton observed, ‘On the basis of the UKUSA agreement, Britain’s outposts of empire thus became as important for Washington for SIGINT collection as they had been for the British government itself.’ Not only was America unable to collect a significant amount of Sigint from home, but also given the enormous size of the emerging Sino-Soviet bloc it was only through a cooperative pooling of resources that the partners could hope to collect worthwhile data streams. Britain’s residual empire and its English-speaking Commonwealth partners thus offered American Sigint the best chance of encircling the USSR and China with listening posts and relay stations.
Intertwined as Hedgerows
The cautious and gradual sharing and collaboration of these early years quickly developed into a Cold War Sigint relationship akin to a ‘complex spider’s web of cooperation.’ And, like a fly in a web, it is very hard to extricate oneself once caught up in it. Indeed, over the course of the Cold War, NSA and GCHQ, the two largest nodes in the partnership, became so close as to be nearly inseparable. With remarkable historical foresight, the UKUSA negotiators predicted, ‘The fruits of the work completed at the conference might not be seen immediately, but no one could doubt that, from a long-term point of view, the recent meetings were an important stage in US-British relations in general and in US-British Communications Intelligence cooperation.’ How right they were. A former MI5 director general averred, ‘The institutional integration that has flowed from the 1946 UKUSA Agreement is so widespread that Sigint customers in both capitals seldom know which country generated either the access or the product itself.’ UKUSA was the impetus to formalise and build upon the already strong foundations.
Remarkably, the signatories of the post-war UKUSA agreement agreed not only to cooperate, but also to refrain from targeting each other’s communications. Indeed, it was almost structurally impossible to do so as a result of the Secret Treaty. As Edward Travis presciently observed during negotiations with the Americans in 1945, ‘Since any withholding would be obvious, it would be very easy for suspicion to be aroused on either side.’ To deal with any potential suspicion, each side settled on the following procedure for justifying withholding: ‘Such exchange will be unrestricted on all work undertaken, except that upon notification of the other party information may be withheld by either party when its special interests so require. Such notification will include a description of the information being withheld, sufficient in the opinion of the withholding party, to convey its significance.’ Moreover, this clause was only to be invoked rarely: ‘It is the intention of each party to limit such exceptions to the absolute minimum.’
Unlike the ad-hoc and personality-driven CIA / Secret Intelligence Service Humint relationship, the US/UK Sigint relationship has been less fractious, and more durable with more permanent fixtures, including expensive infrastructure and outposts, and a clearer division of labour and geography. If Humint tradecraft is ‘more art than science,’ Sigint is ‘more science, less art.’ And, importantly, Sigint is less vulnerable to personality conflicts. NSA officers at the working level maintained a relationship that was only marginally influenced by the interactions of political leaders or very senior intelligence officers. Former GCHQ director Sir David Omand concluded that NSA and GCHQ officers were given a freer hand to manage their relationship because their political masters did not fully understand the technical implications of what their subordinates were doing. In this sense, the liaison relationship could be described as ‘experts calling experts’ because cryptography was a specialist language that few understood.
If the political masters did not quite understand the technical machinery involved, the civil servants were at pains to present the significance of the Sigint alliance, both in terms of the wide-ranging scope and the tremendous scale of the joint collection. The Sigint relationship became more than just synergy; it was true interdependence. ‘No spy’ agreements are one thing, but the Five Eyes partners are so interwoven that they could not disentangle themselves without serious harm to the collection effort even if they wanted to do so. This evolution has been dramatic: in 1915 the American military would have nothing to do with joint Sigint, but by 2013 American NSA contractor Edward Snowden was able to steal tens of thousands of top secret British Sigint documents without setting foot in Cheltenham.
If common networks and professional ties bound the Sigint agencies together, the wider relationships among the Five Eyes countries have not been without friction. The Cold War occasionally saw political disputes intrude on the working level of Sigint collaboration: fears of communist influence in the late 1940s delayed the incorporation of Australia – and, by extension, New Zealand – into sharing arrangements with the Americans; disputes over British policy in the Middle East and Europe led Henry Kissinger to demand a suspension of Sigint exchange with the UK in 1973; and the following decade New Zealand’s decision to refuse port visits by nuclear-powered and nuclear-capable warships saw the Americans respond by limiting the supply of intelligence, including Sigint. These incidents are a reminder that diplomatic controversies can and do pressure liaison arrangements; that the partnership has withstood them is partly because its members developed creative work-arounds and persuaded their leaders to reaffirm Sigint commitments, and partly because the need for burden sharing is reinforced with each new innovation in collection technology.
If the Five Eyes machinery continues to operate without serious disruption, a more fundamental question looming over it is the changing nature of Sigint expertise. Digital communications have transformed the domain and role of these once-secret agencies. Whereas they were originally focused on the technical contest over codes, these agencies are now pivotal actors in cyber and telecommunications decisions. Some national variations – over encryption regulations, data retention, supply chain security, disclosure of threat assessments, or even judicial oversight – can be expected to arise among the Five Eyes countries. It will be the task of Sigint officials in each capital to foster a common view of these and other issues, limiting the prospect for major divergence between their political masters. This was a feature of the debate in 2018-2019 over the supply of Chinese-built components in the UK’s 5G mobile network, which carried a vague threat that intelligence exchanges could cease if the British government failed to adopt security restrictions to the Trump Administration’s liking. While this dispute was resolved, it demonstrated how politics and policy-making is uncomfortable territory for the intelligence professionals who uphold Five Eyes cooperation.
Bonds in peace and war
For every politician inclined to use the Sigint relationships as diplomatic leverage, there are more who seek to appropriate the image of Five Eyes solidarity for other purposes. Working meetings and joint communiques between these countries are now regularly branded as ‘Five Eyes’ initiatives. While these usually have limited impact on Sigint practice, they do suggest that the wider alliance politics of these countries tilts towards more integration and coordination over time, not less. Indeed, the Five Eyes agencies function as military instruments in support of reconnaissance and surveillance, information operations, and other conflict roles. The drive for a networked, cross-domain military force in each country has also necessitated more fusion of Sigint capabilities, and its use on the battlefield. In the cyber and space domains, the capacity of the Five Eyes to monitor, penetrate, and analyse potential adversaries will be a powerful, if not determinative, asset in any future conflict.
For all this synergy, there are compelling reasons for the Sigint agencies to deepen ad hoc liaison with other partners in years to come. As a unique form of partnership, it has nonetheless operated within a broader web of relationships with friendly, if somewhat competitive, Sigint agencies elsewhere. While the idea of second and third-tier allies such as Germany or Japan acceding to the ‘no spy’ club of the Five is a non-starter, a degree of inter-operability in military Sigint and cyber operations will be needed for these and other countries to provide credible assistance for Western strategy in Europe and Asia. Throughout the Cold War the Five Eyes never had to fight a hot war against their principal intelligence adversary, the Soviet Union; in effect, this spared them from the dilemma of how to harness peacetime Sigint for wartime intelligence support, although it raised questions about the viability of NATO’s collective response to Warsaw Pact aggression. Revising the use and circulation of highly protected Sigint to enhance future military readiness – and thus deterrence – among a wider group of countries in the twenty-first century will sit uneasily against their instinctive preference for exclusivity.
If the core of this Sigint alliance remains unchanged and unparalleled, the Five Eyes agencies will nonetheless face new problems of policy-making and diplomacy as the technologies of communications and warfare change. The history of their cooperation is an anchor for stability, reinforced by trust among intelligence professionals and the wider politics of a like-minded, Anglophone community. Cryptanalytic interdependence may indeed have reached a stage of institutionalisation where it is difficult to see how it could come undone without great damage to national Sigint in each country. But maintaining these working arrangements is not sufficient nor guaranteed in perpetuity; their shared knowledge and expertise must be exploited for advantage in new areas of international security. If resources and technology invest the Five Eyes partnership with more strategic influence, it also bears a greater burden than could have been foreseen by its wartime founders. If its history is any guide, the alliance will rise to twenty-first century challenges.
This analysis is the view of the authors and does not represent any official United States Government or Department of Defense position.