CIPS Working Paper No. 14

May 2021

(click  here for French translation)

Srdjan Vucetic, University of Ottawa

Earlier this year the Five Eyes partnership passed another major milestone: 75 years of the UKUSA Agreement on signals intelligence, a.k.a. SIGINT, co-production. To mark the occasion, Jeremy Fleming, director of the UK Government Communications Headquarters (GCHQ), and director of the US National Security Agency (NSA), Gen Paul Nakasone, issued a joint statement. “Together, we are greater than the sum of our parts,” it said, with “we” extended to Australia, Canada, and New Zealand as well. 

The statement speaks of an “extraordinary partnership” as well as an “alliance.” The first term is entirely suitable in this case, as such a long-standing, trusted collaboration among five sovereign countries can only be described as extraordinary. The word alliance has a more elevated normative connotation, which in turn makes government PR teams nervous. “British officials point out that the Five Eyes is not an alliance — its remit does not go beyond intelligence,” notes one Financial Times columnist. Most of their counterparts in Canberra, Ottawa, Wellington, and Washington would no doubt agree. For instance, the publications and press releases of Canada’s security and intelligence organizations, the Communications Security Establishment (CSE) and the Canadian Security Intelligence Service (CSIS), suggest a clear preference for “partnership” and its derivates – “alliance partners” or “allied partnering arrangement,” for example.

Politicians, pundits, and journalists follow different institutional and discursive rituals, hence their vocabulary revolves around not only “alliance” and “partnership,” but also “bloc,” “club,” “coalition,” and “network,” all which are in turn collocated with “friends,” “community,” “arrangement,” and, in social media communication at least, “FVEY,” the practitioners’ own acronym. We likewise learn that the term Five Eyes can be used in both plural and singular form, as a proper noun (Five Eyes as opposed to the Five Eyes) as well as an adjective. Looking further afield to languages other than English, we see an even larger universe of phrasing choices – especially now that Beijing has officially threatened to “blind the Five Eyes,” whereas officials in Tokyo are saying that the time has come for formally adding more eyes to the “spy alliance.”

Why are so many terms used to describe (the) Five Eyes today? Part of the answer lies in a fact that Australia, Canada, New Zealand, the UK and the US have come to share many extraordinary five-way partnerships, not “just” the spy alliance. The intensification of inter-governmental ties has been nothing short of remarkable — look no further than various joint statements on encryption, Hong Kong, or Nagorno-Karabakh. Thus, when Nanaia Mahuta, the new New Zealand foreign minister spoke at a joint press conference with her Australian counterpart Marise Payne on April 22, 2021, she felt strongly compelled to clarify that the government in Wellington was against such “expansion,” including when it comes to the China policy: “The Five Eyes arrangement is about a security and intelligence framework. It’s not necessary, all the time on every issue, to invoke Five Eyes as your first port of call in terms of creating a coalition of support around particular issues in the human rights space.”

In addition to policy expansion, we have also seen an expansion in transparency and knowledge. Compared to even five years ago, the media and therefore the citizens know more about the nature of “Five Eyes work” done by spies, soldiers, and diplomats. The two expansions are indeed connected.

In this working paper I engage a handful of International Relations (IR) concepts that I think can both illuminate and complicate our understanding of the Five Eyes as a historical-sociological phenomenon – its distinctive features, its multiple forms, and its temporal and spatial coordinates.  Suffice it to say, my purpose in doing so is neither to “correct” the official/officious labelling practices nor to somehow “reconcile” them with scholarly conceptual taxonomies. Rather, it is to invite further reflection on an object of shared interest. Perhaps such exercises matter only to those who study the Five Eyes, and not especially to those who are studied. Still, there is no denying that words we use matter, and that basic linguistic and conceptual reflexivity can serve to advance knowledge, scholarly or otherwise. This, in turn, can provide some relevant input into policymakers’ strategic thinking and thus into the formulation and implementation of future policies.[i]


If Not Alliance, What?

Let us start with a codebook definition of alliances by Brett Ashley Leeds: “An alliance is formal agreement among independent states to cooperate militarily in the face of potential or realized military conflict.” This is emphatically not the kind of alliance NSA and GCHQ directors had in mind. The document signed in Washington on 5 March 1946 – dubbed BRUSA, and then UKUSA Agreement – does not mention military cooperation or military conflict once, and neither do the subsequent iterations of this document, which Australia, Canada and New Zealand signed in the following years as they each formally joined the alliance as “equal partners.” (Some of these documents have now been declassified and released to the public, allowing historians to compare between different versions.)

UKUSA is not your “typical” interstate treaty either. As historian John Ferris observes, “The UKUSA agreement is a formal arrangement, but not a national treaty. It is deniable, non-binding, and open to instant abolition. In the strictest sense, UKUSA is merely a memorandum of understanding between two national interdepartmental intelligence coordinating groups, from which broader multinational arrangements have emerged.” 

It is these broader arrangements that make defining the Five Eyes so challenging today. Let us consider media coverage first: Figure 1 shows a count of the relative frequency of the term “Five Eyes” appearing in articles published in the world’s top English and French language newspapers from 2010 to 2020. To begin with, the partnership’s newsworthiness is relatively recent. In the 2012, the corpus contains only one discrete newspaper item – an op-ed on Australia’s intelligence community published in The Sydney Morning Herald. Eight years later, the same newspaper published 74 such items, which per to the same measure puts it behind The Guardian (75), Global Times (86), South China Morning Post (112), and The Australian (169). We also see significant increases in instances of the term appearing in 2013, in 2017, in 2019, and in 2020. The first jump corresponds to the disclosures former NSA sub-contractor Edward Snowden made that year about NSA and its Five Eyes partners. (The impact of the affair cannot be stressed enough: in addition to giving incredible insights into contemporary surveillance and international security cooperation, Snowden prompted vigorous debates on the effects of the US-led “war on terror” on executive power and civil liberties.)

The increases of mentions after 2017 index the turbulence of the Trump years as well as the challenges presented by a rising, more assertive China. As Washington moved away from a policy of “engagement” with Beijing and towards that of “strategic competition,” there came both pressure and opportunity for US allies to recalibrate their foreign policies. This is the context for Mahuta’s statement above; in contrast to her government, the governments of Australia, Canada and the UK have been comfortable with moves to “diplomatize” and “politicize” the Five Eyes. The newspapers followed suit. In the earlier years, they treated “alliance” and “partnership” as interchangeable, sometimes alongside the terms “bloc,” “club” and/or “network,” albeit strictly in the context of intelligence. Then those same terms migrated to articles on diplomacy. Moreover, this appears to be the case across the newspapers sample, regardless of country of publication, language, and perceived ideological leanings.[ii]


Figure 1: Mentions of “Five Eyes” in 34 English & French Language Newspapers Worldwide, 2010-2020

Srdjan Vucetic. 5 Eyes. Graph.png


On one level, the observed terminological shifts reflect the changing scope of Five Eyes cooperation from a spy alliance to a more ambitious alliance. But if we accept a theory that language (re)produces reality, we might then say that terminological and institutional changes are co-constituted.  Thus, when historian Wesley Wark labels the present-day Five Eyes a “grand political coalition…G5,” he offers a linguistic means with a potential to both reflect and reinforce the new institutional and political reality.

With this important point in mind, what might IR scholars call the Five Eyes today? Leeds’ codebook offers a clue: “any states that share policy positions and coordinate their actions might be called aligned.”  Following Thomas S. Wilkins, Vidya Nadkarni, Coleen Chidley, and H. D. P. Envall and Ian Hall, we know that “alignment” is best seen as an umbrella concept that subsumes other concepts. In the sections that follow I will reflect on four such “sub-concepts” in turn: “strategic partnership”, “security community,” “concert,” and “coalition.” While each, I suggest, is useful for thinking and talking about the Five Eyes, their all-too-narrow focus on interstate relations also obscures as much as it illuminates. For this reason, I turn to the fifth concept, “network.” The advantage of this concept lies in the way it challenges the artificial division between “domestic” and “international,” allowing us to trace events and relations that circulate within, across, among, and around the five states that constitute the Five Eyes today.  


Strategic Partnership

In IR, the word partnership comes with adjectives. “Strategic partnership” in particular entered the field’s lexicon in the 1990s, partly as appropriation of the label the then Chinese and Russian governments used to talk about their own alignment efforts. Going by Jonathan Holslag’s five-part definition, the Five Eyes is a strategic partnership par excellence. Common interests are present, as are long-term expectations of common interests, pursuit of joint cooperative ventures across multiple issues areas, global reach, and recognition of the partnership’s necessity and specificity. Envall and Hall’s definition also applies, according to which the Five Eyes would be seen as a constellation of practices – patterned embodied actions – in “security.”

I will come back to practices in a minute. For now, let us only say that strategic partnership does not necessarily signal an affinity of interest or mutual benefits that FVEY officials have in mind when they talk about “extraordinary partnership.” “Some partnerships link friends or potential friends; some link actual or putative rivals,” write Envall and Hall. We might say that Britain and the US were both – rivals who became friends. As I explain next, this history is crucial for understanding the Five Eyes in whatever form.


Security Community

Compared to strategic partnership, this concept has a much longer lineage in IR theory.  According to most definitions, security communities are institutionally integrated, transnationally interdependent groupings of states within which inter-state conflict is “unthinkable.” Security communities are also said to share common values, mutual sympathy, trust, and a sense of “we-ness.”

The Five Eyes arrangement meets these criteria. In addition to the extensive and interlocking practices and institutions of cooperation at the inter-governmental level, this entity is said to be characterized by some commonalities at the transnational level. Five Eyes officials say so themselves, variously referencing common (read: U.S.) intelligence and security standards, common heritage, common language, and the like. We see this in the aforementioned joint statement, for example: “Our alliance…,” GCHQ and NSA directors state, “[is] built on a history of strong shared values, including respect for privacy and the rule of law.”

While the majority of accounts date the birth of the Five Eyes to the 1940s and 1950s, the joint statement says it “began before World War II.” In light of the aforementioned concert behaviour at Versailles, this is an eminently defensible interpretation. Not only were all five countries allies in World War I, but US and British intelligence agencies also first learned to work together during this conflict – see, for example, cooperation between the intelligence staffs of the British and American Expeditionary Forces in France. On top of this, as David Haglund has recently argued, the Great War brought about a precipitous and lasting decline of Anglophobia in Washington politics, thus opening up new alignment possibilities down the road.

Following Coral Bell’s 1964 book, one might even argue that the beginnings of the UK-US alliance can be tracked back to the turn of the twentieth century. As British power declined relative to that of Germany and the United States, successive governments in London decided it made sense to oppose the former by means of a strategic partnership with the latter. Driven by cold realpolitik and ideas and affects of “Anglo-Saxon” racial solidarity, the strategy in turn led to what IR scholars call “hegemonic transition” – the peaceful replacement of Pax Britannica with Pax Americana.

Enshrouded in popular myths and elite ideology, the Anglo-American “succession,” as this transition is also dubbed, is one of the most consequential transformations of our time and a major point of origin of what some IR scholars refer to as “the liberal international order.” Indeed, as Washington moved to claim the mantle of world leadership at the war’s end, the British did not protest. On the contrary: on the same day that UKUSA was signed, Winston Churchill was in Fulton, Missouri, arguing that a “special relationship” with Britain was a fundamental American interest. Many in the audience were sceptical, but the expansion of Soviet power in Europe and the communist victory in China soon proved the former Prime Minister right. A State Department policy paper from April 1950 presaged: “To achieve our foreign policy objectives we must have the cooperation of our allies and friends. The British and with them the rest of the Commonwealth, particularly the older dominions, are our most reliable allies, with whom a special relationship should exist.” 



The next alignment concept on our list is “concert,” a term usually associated with past efforts by empire-states to organize a system or systems of conferences to deal with issues of common concerns. As Jennifer Mitzen explains in her path-breaking study of the Concert of Europe, the order that the major European powers in response to the upheavals of the Napoleonic War restructured international relations in ways that had profound consequences for the subsequent course of history. The key mechanism in this process was face-to-face diplomacy, which she argues transformed state behaviour by creating a collective identity as well as by mobilizing “international public power of governance.”

Although the Five Eyes states do engage in it, concert diplomacy is still a relatively recent practice; ten years ago the foreign ministers of the five states in question would never self-identify as a Five Eyes grouping, let alone release any joint statements to the public. Viewed from more broadly – as well as more historically –, however, the international public power governance the Five Eyes generated by acting together – either by themselves or in concert with other states – is at least a century old. A case in point is the post-World War I peace conference at Versailles, where the delegations of what we now call the Five Eyes nations plus South Africa acted in concert, mainly on Australia’s behalf, to veto the Japanese delegation’s proposal to insert a “racial equality” clause into the League of Nations Covenant.

While such “civilizational” logics would in due course be banished from diplomatic discourse and practice, their infusion into the foundations of the so-called liberal international order remains relevant for contemporary international processes. A case in point being the Five Eyes’ ongoing geopolitical and ideological feud with Beijing, which some IR scholars treat as a testing ground for theories of “the politics of recognition” and of “norm contestation.” Echoing its attempts to cast the Cold War as a racial war, the latter, represented by various Chinese Communist Party voices, are now increasingly willing to publicly refer to theFive Eyes in retrograde “civilizational” terms. These are rhetorical power moves whose purpose is to undermine and eventually replace some of the basic norms of said liberal international order.

On both past and current trends, Beijing’s moves are unlikely to be successful. For one thing, instead of projecting confidence and attracting followers, China’s essentialist rhetoric is being almost universally interpreted as a crude attempt to deflect attention from troubling developments at home. For another thing, Chinese accusations tend to strengthen the resolve of some Five Eyes diplomats to rally international democratic support for concert-like action against Beijing, while making it harder for other Five Eyes diplomats to opt out of joining joint Five Eyes declarations against China’s human rights abuses.



Unlike the terms security community and concert, “coalition” is part and parcel of the policy lexicon today. As one recently published New Zealand foreign affairs document explains, foreign policies are now increasingly conducted with “coalitions of interests” in mind: “We will also continue to derive benefit from leading and joining coalitions of interest with countries with which we share (to varying degrees) a set of common interests and values, including a commitment to multilateralism, free trade, the rule of law and liberal democratic values.”

Growing interest in the issue-specific, “plurilateral” initiatives that bring “like-minded” states together has much to do with the ongoing transition from the aforementioned United States-led liberal international order to a more fragmented, multipolar world marked by vigorous disputes over the “rules of the game.” Of course, some IR theories would say that the philosophy behind interest-based coalition-making is both ancient and sound. Indeed, it was in the run-up to the Iraq War, the putative zenith of the liberal international order, that then U.S. Defence Secretary Donald Rumsfeld offered his classic realpolitik quip that the mission should define the coalition, not the other way around.

That said, history would suggest that Washington has good reasons to count on its Five Eyes allies to form a seemingly ad hoc coalition in support of U.S. foreign policy goals. One explanation is this outcome is a by-product of a broader “ecosystem”behind the United States-led international order, which resembles a “hub-and-spokes” system with multiple “security triangles” and “quasi alliances” . As “spokes” to the United States’ “hub,” Canada and Australia need not be in a formal alliance to regularly cooperate with each other militarily, in peacetime and international crises alike.

Other structural dynamics feed into this behaviour as well. Looking at the recent history of US, UK and Australian foreign policies towards West Asia, Jack Holland has observed a great deal of “like-mindedness” in how Washington, London and Canberra define threats and opportunities. Admittedly, democratic political leaders always legitimate their foreign policy choices by crafting discourses for specific national audiences. However, what is remarkable about the US, UK and Australia is that their discourses tend to invoke the same “Anglosphere” core ideals and interest. Viewed as such, US, UK and Australian foreign policies are at once national and embedded, via processes of co-constitution and mutual influence, in a “transnational political space.”

Following Holland’s analysis, the Anglosphere is thus a security community that also manifests itself as a two-speed military coalition, with the US, UK and Australia – what he calls the “old Anglosphere coalition” – exhibiting greater willingness to go to war than Canada and New Zealand. The same perspective may well be useful for understanding Five Eyes behaviour in some other domains as well, a contemporary case in point being the “fissures” on Huawei and China more generally.    

In Holland’s analysis, the conceptual heavy lifting goes to language, discourse, and identity, and specifically to state- and nation-level ideas about who “we” are, were, or want to become. However, the author is clear that ideas do not float freely but are rather anchored in material structures, organizations, and everyday practices. This is a crucial point for we cannot fully understand the Anglosphere without due attention to the newsmedia, finance, think tanks, and academia (IR included). Add to this the longevity of some of these anchors  – Churchill’s term for the Anglosphere was the “English-speaking world” – and we might better appreciate why the Five Eyes countries so often end up seeing eye to eye, thus jointly generating what we might call “Anglobal governance.

Whatever their view of the Anglosphere, most scholars would agree that Five Eyes coalition behaviour has been path dependent. The circumstances of World War II set things in motion, and then, for seven decades, member states kept practicing what they have practiced before – because continuity usually carries more benefits than change, and sometimes also because change is literally unthinkable. Disruptions of course occurred many times, but the overall arrangement has proven robust so far. This is important to keep in mind as we ponder the recent diplomatic and media tumult over the New Zealand government’s desire to curb Five Eyes expansion(-ism). Intelligence co-production will continue, and with it, a degree of policy coordination in other areas as well.

Path dependence alone cannot shed much light on the ongoing transformation of the Five Eyes from a “mere” SIGINT alliance into an entity approximating an actual alliance. For a more complete picture we would need to bring IR perspectives on alignment together with the large and sparsely connected literatures on the “internationalization of the state,” “policy spillover effects,” “security problematizations,” and “transnationalization of defence,” to name but some. Note as well that path dependence does imply that friends will always be friends, meaning that tactical- and operational-level Five Eyes cooperation will always muddle through political-level crises. (A good antidote to such teleological fallacies is Bill Robinson’s exploration of a hypothetical scenario in which the US leaves FVEY).

Like the previous three, the concept of coalition can prompt the right reflections about the history and character of the Five Eyes. However, alignment concepts are best tailored for those interested in interconnections between states as corporate actors. Those more interested in bringing the international and the transnational together would likely begin with alternative concepts. One such concept is “network.”



Looking specifically at intelligence, the Five Eyes is a “community of practice” made up of hundreds of individuals who generate international cooperation by work together in and through networks. The closest-knit network is in SIGINT, which the GCHQ-NSA document defines as intelligence produced through intercepting “communications, translation, analysis and code breaking information.” There, intelligence is not shared so much as co-produced, with assets from the participating countries working so closely together as to be nationally indistinguishable.

Following Table 1, however, the Five Eyes is not “just” a SIGINT network. Rather, it is a “network of networks” of at least twenty-two intelligence agencies. I say “at least” because routinized quintuples have emerged between many other “intelligence” entities as well – national assessment bodies, national counter-terrorism centers, national cyber security centers, national police forces and more.


* Table 1 is partially derived from James Cox, Brig-Gen (Ret’d), “Canada and the Five Eyes Intelligence Community,” Canadian Defence and Foreign Affairs Institute Strategic Studies Working Group Paper, December 2012, p. 10. 


Seen this way, no single club of three, four and five-letter acronymed agencies drives the spy alliance as such. Rather, we have a combination of mechanisms and practices operating across many “nodes.” Some of these are based on formal arrangements and discrete written agreements, others on assorted social processes that transcend formal boundaries, and still others, namely the SIGINT network, function as a de facto single agency. Furthermore, if we were to try and visualize this network as such, we would also see that nodes therein do not coordinate with each other. Indeed, we can easily imagine a scenario in which we have two separate Five Eyes events taking place one the same day, in the same state, within thirty kilometers of each other – one at the CIA headquarters in Langley, one at the Drug Enforcement Administration (DEA) in Chantilly.  Finally, as both Wark and Steven Loleski have recently observed, there is no reason to assume that the newer networking processes will always be operationally or politically complementary with the established activities. 

In sociological approaches, “intelligence” is never limited to intelligence agencies. Rather, it a broader, heterogenous social space where spies operate alongside activists, corporate executives, journalists, lawyers, and scientists, among other groups of state and non-state actors. As Ben Jaffel, Hoffmann, Larsson, and Kearns have recently argued, these approaches are essential because they allow us to better appreciate not only the scope, depth and scale of transnational intelligence networks and practices, but also the conditions under which they maintain and collapse the legal boundaries between citizens and non-citizens. Prima facie, this chimes with some personal(ized) histories of the Five Eyes, where emphasis tends to move away from “inter-state relations” and toward “micro-level” mechanisms and processes that create networks of exchange within and across the member states.

Network-theoretic approaches can also help us make sense of a proliferation of Five Eyes networks in many non-intelligence policy domains. Using interview and open source data, Tim Legrand has identified close to forty such “Anglosphere policy networks,” as he labels them. Carrying names such as “Five Nations Passport Group” or the Five Eyes’ defence ministers conferences and linking state bureaucracies without the intercession of formal diplomatic processes, these transgovernmental, networks are informal and uncoordinated. (Again, the meetings of “Five Nations Passport Group” have little to do with the Five Eyes’ defence ministers conferences, let alone with the equivalent meetings between the agencies appearing in Table 1.)  Nevertheless, as Legrand shows, they are important sites of direct and indirect policy transfers, and, in some cases, of “intermestic” governance.

Five Eyes networking practices exist in the military realm, too. The abbreviation “ABCANZ” refers to an international program that promotes land force interoperability and standardization – “human, procedural and technical” – among the armies of the Five Eyes countries. The first to come together, in 1947, where the American, British, and Canadian armies – the “ABC armies.” In 1963, the Australian army joined, adding the second “A” to the program’s name. The New Zealand army, an observer since 1964, became a full-fledged member in 2006. Similar networking practices exist in the other domains – maritime, air, space, cyber, and special forces. On top of this come other practices: those that allow military professionals and citizens of one Five Eyes country to join the armed forces of another with relative ease. Furthermore, cooperative arrangements between the militaries and defence bureaucracies of the Five Eyes counties are cemented through cooperation at the level of defence industries as well as through exclusive defence trade cooperation treaties that remove some restrictions on some military goods transfers. Finally, keep in mind as well that the reigning defence “white papers” in Australia, Canada, New Zealand and the UK all duly mention the Five Eyes and the benefits this “unique and highly valued” partnership-cum-alliance  – the labels used once again vary – brings to each member nation in terms of information-gathering, cost-saving and burden sharing.

Network perspectives complicate conventional wisdom about world politics because they can focus on relational space between either states or people – or, indeed, both. As such, they are well-positioned to tell us something heretofore new about the diversity of actors and processes involved in the production of the Five Eyes.



The character of the Five Eyes is changing, and so is its lexicology. What I have provided here is a brief discussion of five IR concepts that I believe can help us gain a better understanding of these changes. A simple but unsatisfying conclusion to draw from it is that the Five Eyes requires multiple conceptual lenses. However, network seems especially promising because it sits between the dichotomy between “domestic” and “international.” As such it can offer a richer set of reflections on the overlapping, nested, and ever-changing social interaction that make the Five Eyes possible. A more provocative conclusion, then, is that the extraordinary partnership in SIGINT is but one node in a much broader constellation of Five Eyes networking practices across time and space. Following the co-constitutive logic of “domestic” and “international,” we furthermore see that while most of these networking practices are informal, uncoordinated, and transnational, some are in fact a function of written interstate treaties signed decades ago. Individually and cumulatively, however, these networking practice provide political and legal basis for a remarkable communitarization of policymaking that we are only beginning to comprehend.

To say that multiple conceptual frameworks are needed is simultaneously a call for further conceptual analysis. Our discussion of security communities, for one, has already brought forth concepts of regionalization, hegemony, empire, and assemblages which in turn trigger larger questions about international order. By the same token, we cannot fully discuss Five Eyes networking practices without due attention to Jason Dittmer’s “counter-history” in which UKUSA emerges as a messy objectof socio-technical relations so complex that we can only comprehend it by transcending our conceptual sphere altogether. Whatever one’s starting point, however, should not forgo opportunities to be reflexive the semantic, social, and politics orders that their own concepts (re)produce.



[i] For excellent research assistance, I remain grateful to Caroline Dunton and Mana Khosrowshahi. I likewise remain grateful to the participants of the 2020 Center of International Policy Studies (CIPS) Twitter Conference, “Understanding the Five Eyes.” The proceedings of this conferences are available on the CIPS Blog.

[ii] To ensure that the findings is not some sort of artefact of global newspaper culture, I situated this finding against the equivalent data from the weekly newsmagazine Economist as well as from a selection of think tank and parliamentary documents. For think tanks I read 19 reports, working papers, blogs, and similar materials published between 2012 and 2019, while for parliamentary discourse I relied on quick keyword searches of spoken and written contributions in the five national legislatures. The results stayed the same: with Snowden, “Five Eyes” becomes worthy of attention as a spy alliance; then, with Trump, as something bigger than a spy alliance.