Existential security: Safeguarding humanity or globalising power?

Nathan Sears' (2020) exploration of how a policy of ‘existential security’ might be fostered represents one of the first efforts to systematically think through security and how it might relate to thinking about existential risks. The concept of existential risk emerged in the early twenty-first century (see, e.g. Bostrom, 2002). It refers to the idea that there are a class of hazards which may ‘threaten the premature extinction of Earth-originating intelligent life or the permanent and drastic destruction of its potential for desirable future development’ (Bostrom, 2013, p. 15). As a new field of study, Existential Risk Studies (ERS) is small but quickly expanding. A number of research centres have attracted significant attention – from both the media and policymakers – and large amounts of funding from high-profile private individuals and philanthropic foundations. The concept is also entering mainstream political discourse. In recent months, UK Prime Minister Rishi Sunak has met with leaders of industry in AI research to discuss existential risks1 and references to their existence have proliferated in both the vernacular of political elites and in policy reports and white papers. But how all this relates to the concept and practices of security is by no means straightforward. Work in ERS is oriented around two core concepts: risk and security. Yet before Sears' interventions, there had been little work that engaged in a sustained and reflective manner on the meaning or function of either of its core concepts let alone the relationship between them. Security is traditionally a concept attached to the state, whereas risk has its origins in insurance and the private sector. As they have grown gradually closer, both concepts have been further challenged by radical uncertainties and potential catastrophes linked to globalisation and technological advances (Petersen, 2016). Early grapplings with catastrophe, globality and technology can be found in the ‘nuclear realism’ of post-World War 2 nuclear age security thinkers like Bertrand Russel, Günther Anders, John Herz and Lewis Mumford concerned with geopolitics and the survival of human civilisation (van Munster & Sylvest, 2016). Daniel Deudney pointed to ‘planetary geopolitics’ generated by global machine civilisation (2018), and Copenhagen School writers have recently explored ‘macro-securitisations’ – some of which potentially frame the planet or the whole of humanity as the valued referent object (Buzan and Wæver, 2009). But Sears' work broke new ground in its direct appeal to and embrace of globalised security logics and theory. This helpfully sets up a space for working further through the relationship between security and anthropogenic existential risks and their governance. In ERS, many have proceeded from the starting point that these are ontologically stable categories – that risks, while often hard to assess accurately, are obdurate facts and that security is (in principle) an attainable condition and a normative good. Indeed, definitional accounts in the field have generally gone little further than emphasising the distinctive scale of risk that they are concerned with (Bostrom, 2002, 2013) or have instead focused on probabilistic accounts of the ‘chance of a terrible event occurring [and] wiping out intelligent life’ (Cotton-Barratt & Ord, 2015). Engagements with security have been characterised by what we might term a common-sense embrace of the ethic of protection implied by ‘securing’ something, or by an implicit preference for the maintenance of order (Bostrom, 2019) over the messy possibilities created by the global interactions of people, things, and knowledge. ERS scholars have thus taken an instrumental approach to the concept of security, advocating that an increasing range of possible risks to humanity's future—including knowledge dissemination (Seger, 2022) and scientific research (Bostrom, 2019)—ought to be secured against. The ‘Vulnerable World Hypothesis’ (Bostrom, 2019) represents the most dramatic indication to date of where such an uncritical embrace of security logics and practices may lead, namely to pervasive planetary surveillance and pre-emptive global micro-policing. In large part, this objectivism is a consequence of the epistemic and disciplinary history of the field of ERS, with its roots in Enlightenment traditions of positivism and rationalism, and its relative non-engagement with sociology, social theory and critical analyses of security. Nevertheless, ERS is also self-avowedly committed to interdisciplinary approaches and dialogue (Beard & Torres, 2021). Nathan Sears' scholarship took up this invitation from the perspective of International Relations and, more particularly, Security Studies. The discussion we venture to here reflects on Sears' legacy and takes his contribution as a starting point for thinking further about the relationship between security and the study of existential risks. Sears' ‘Existential Security: Towards a Security Framework for the Survival of Humanity’ (2020, hereafter referred to as Existential Security) is the first account within ERS of how scholars in the field might explicitly define ‘security’ and translate it into a framework for motivating policy choices geared towards existential risks. His essay identifies and critiques two extant frameworks for security policy—human security and national security—contrasting them in terms of scale, referent object, threat prioritisation and means of enactment. He also reflects on a number of competing definitions of what security is and how it relates to other aspects of politics and human values, constructing an account of what security is and what its political status might allow and legitimate in approaches to policy. Ultimately, he presents a largely positive picture of both the attainability and desirability of security and of the utility of deploying security as a framework to elevate the importance and urgency of existential risks in contemporary political decision-making. Where Sears differs from other ERS treatments of ‘security’ as a black-box, left implicit or advanced as a self-evident normative good, is when he brings an explicit theory of security to bear on the field, namely Securitisation Theory. Taking the Copenhagen School security ‘grammar’ of existential threats to valued objects legitimating exceptional means, he makes sense of existential security in terms of existential anthropogenic threats to humanity (or civilisation) and demands of politics and policy that identify the exceptional measures necessary to secure that object. Existential Security is therefore an effort to translate the existing logic of security into a framework that is appropriate to the domain of existential and catastrophic risks. On the one hand, we might follow Sears in viewing this as a more or less straightforward process, amending the referent objects, prioritisations, and proposed mechanisms in order to gear security practices towards the lofty goals of planetary safety or species survival. On the other hand, the translation of ‘security’ into a radically new setting and context requires at least pausing to question such a move. How security is re-negotiated in reference to new problems and in new contexts has recently been noticed (by Copenhagen School authors themselves) as a major part of security politics itself (Berling et al., 2022). What follows is a closer look at three core moves involved in what could be summarised as an attempt at securitising existential risk. Existential Security follows prominent thinkers in ERS (Bostrom, 2013; Ord, 2020) in proposing that ‘humankind’ or ‘civilisation’ is increasingly existentially threatened, largely as a result of ‘our’ Pandoran relationships with technological systems we cannot comprehend, or our Icarian hubris in developing artefacts that exceed our capacity to control them. Sears correctly noted that ‘security literature currently lacks a theoretical and policy framework for existential threats to humanity’ (2020, p. 255), yet the first step of this securitising move—the identification of ‘humanity’ or civilisation as the object to be secured—involves its own set of risks. Adopting humanity as a referent o bject potentially belies a multiplicity of entities or identities; as a process, it should not be done without engaging with the history of contestations concerning where the boundaries of humanity—let alone civilisation—lie (Barnett, 2018). A singular notion of humanity risks homogenising all humans (Agier, 2010) though the question of who qualifies as ‘human’ in the sense of being distinct from nature has historically excluded certain people, grouped instead with nature as ‘savages’ etc. (see Patel & Moore, 2017). Those included are articulated as one vulnerable biological population and potentially one species-agent of history. Protecting a singular humanity or civilisation—increasingly a feature of debates about global challenges and risks—glosses over and potentially (re)produces inequities and asymmetries, both in terms of exposure to extreme risks and their production (Cremer & Kemp, 2021). Aggregation of the world's human population—both present and potentially also an unspecified number of future generations—into the biological species category of humanity, or the cultural entity of civilisation within ERS, is, therefore, a much more performative move than many scholars in the field realise. In addition, to place a certain category of risk in the position previously held by threats in security discourse—as that which renders the referent object in need of emergency assistance and protection—also requires scrutiny. The tendency within both the Existential Security framework and the field more broadly, to gloss over differences between ‘risks’ and ‘threats’ elides the different conceptual histories and connotations they carry. Where a threat provides the urgency to security in its direct and ‘looming’ nature, risks can be uncertain and distant. A threat implies a direct causal chain of harm, while risk pertains more to second-order conditions of possibility for future harm (Corry, 2012). Risk thereby broadens the scope of ‘security’ to a more pervasive politics of unease and a proliferation of security measures and generalisation of the politics of the exception (Aradau & van Munster, 2011). Cremer and Kemp (2021) have drawn attention to the consequences of this conflation of risk and security in Bostrom's (2019) ‘Vulnerable World Hypothesis’—with Bostrom advocating for the urgency of extensive exceptional measures in order to secure against the very conditions of possibility of an existential threat being realised. To roll preparedness against potential risks into the politics of security may well aid preparedness against systemic or distant risks but could also facilitate a deeper and wider securitisation than anyone would originally have wished for. With humanity or civilisation as the referent object and risks being a new threat, Sears effectively pointed to an emerging politics of total risk-security that potentially becomes all-pervasive. This should prompt a questioning of the politics of securitisation of this scale, at the very least asking what is—and who defines—humanity, and how are uncertainty and ambiguity translated into security logics? The framework proposed by Sears borrows the ‘grammar’ of security as it evolved within a national and international security framework, as a part of the reason of state, politics of necessity and exception (and a Schmittian notion of politics), transposing it onto a ‘planetary’ level. Doing so he assumed the logic or grammar of security is unchanging despite fairly radical shifts in context: ‘Ultimately, the theoretical logic of security as protection from threats to the survival of some referent object is unchanged by variation in its empirical content’ (2020, p. 257). But is it? The origin of Securitisation Theory in conceptual history makes it doubtful whether the grammar of security can be neatly changed from sovereign to existential security. Although articulated as a formal model or ‘framework for analysis’ (Buzan et al., 1997), for Ole Wæver, the grammar of security is historically produced, albeit evolving with great inertia. Securitisation theory is explicitly just a model (simplification) that takes its point of departure in a dominant emanation of security (in the 1990s). The point was precisely to critique broader notions of security favoured by critical scholars by taking seriously how the concept currently (or then) functioned (Wæver, 1995) namely in terms of post-World War II discourses of national sovereignty and the emerging Cold War context of competing societal models (not just clashing militaries). The shift from ministries of ‘war’ to institutions of ‘national security’ from around 1940—particularly in the newly hegemonic United States—provided the context for Wæver's critique of those scholars who argued security ought to be about non-military matters including environment, poverty, or gender equality, without considering what the security-language game and its ‘us-and-them’ logics would do to the politics of those fields. We might therefore argue, on the one hand, that the notion of existential security overestimates once again how plastic the concept of security really is—assuming not just that it ought to be disconnected from the national security frame, but that it necessarily or already is, perhaps due to the functional need to deal better with global risks. The danger here is a grafting of what is essentially national security practices onto a planetary context. On the other hand, Sears' proposition of an existential security framework, by adopting the existing grammar of security, helpfully poses the question: what form should security discourse and practice take in an age where catastrophic hazards are seen to proliferate, and where the impacts of these hazards will necessarily transcend the boundaries of the nation-state. How dependent is security as a practice on institutions and discourses of sovereignty, raison d'etat and exceptional measures? In the worst case, declaring security a planetary goal smuggles in an attendant global state ontology. Sears' proposition therefore forces us to ask: is national security really obsolete, or do we only wish it to be so, in relation to existential risks? Essentially, what existential security does is not simply an innocent recognition of a growing mismatch between the ‘means of destruction’ (planetary scale) and the ‘modes of protection’ (national security) that needs correcting in favour of the former. It is itself a securitising move that constructs a universal humanity or civilisation in need of security protections. The question is then: will existential security bring necessary emergency measures of a collective kind to bear on emerging catastrophic global threats, or erode ‘normal’ politics of domestic and international society (to the extent that these exist) and potentially legitimate a pursuit, not of global interests, but of a hegemonic set of interests posing as humanity? Again, we feel the Schmittian lesson ought to be learned here: in both defining humanity as an object of governance and in deciding the means through which it might be secured globally, the sovereign is promoted as that which can do the choosing. Liberal order has already been marked not by a rejection of war, violence, and interventions, but by these being justified in terms of securing life and the human (Dillon and Reid, 2009). In simpler terms, there is a danger that existential security will be the practices of a global security elite performed within, and as an extension of, existing international hierarchies. Others have argued that a common humanity could be located as a pre-political space from which the unity needed to tackle global threats could be located—an ‘epochal consciousness’ as Dipesh Chakrabarty (2015) calls it. This is (perhaps) desirable, but any notion of collective global interest is inevitably already shot through with particular (geo)political positions and interests. The persistence of ‘the international’—the division of the social world into multiple uneven units (Rosenberg, 2006)—means that any universal category (of human or civilisation) will be partial or lodged in partial political communities. Legacies of violence and extinction perpetrated in the name of humanity and civilisation make for a bad track record. Added to the statist baggage of existing security practices and discourses, the potential violence of enacting security measures in the name of protecting a planetary or species category should therefore not be overlooked. Sears' work is helpful both in making explicit much of what has otherwise remained implied or assumed concerning security in the field of ERS and in terms of his existential security framework challenging established notions of security. It prompts us to consider the boundaries—political, geographical and conceptual—of security discourse and practice in the face of catastrophic hazards and threats of planetary scale. The challenge for scholars in the fields of ERS and Security Studies is now to ask how emerging planetary challenges and disruptive technologies play into existing frameworks of security, how security might be changing, and ultimately, how the persistence of the international as historically uneven, unequal, and multiple can be reconciled with a normative aspiration for the valuing and nurturing of all lives, past, present and even future. No conflicts of interest. Tom Hobson is a Postdoctoral Research Associate at the Centre for the Study of Existential Risk, University of Cambridge. Olaf Corry is a Professor of Global Security Challenges in the School of Politics and International Studies at the University of Leeds. Data sharing is not applicable to this article as no new data were created or analyzed in this study.