Interview – Swati Srivastava

I was labeled a constructivist before I knew what it meant. At first, Alex Wendt became a guiding light for my thinking. If I had not encountered his work when I did, I would have left graduate school. As I progressed, Ian Hurd modelled how to read texts closely and find value succinctly. Other early inspirations are too many to list here, but I read broadly in the social sciences and humanities from anthropologists of the state to sociologists of law to classics in American political development and literary criticism.

I also began sketching my own way to think about IR inspired by Weber’s view of interpretive social science (informed by Patrick Jackson specifically and ISA-Northeast generally). I knew it was important to produce original theoretical contributions, but I also insisted on original empirical contributions as well. For me, this has meant generating high-quality, systematic, data about global politics, often on understudied cases. I call this a “world-building” approach to research.

In your 2022 book, you write that IR scholars often do not define sovereignty, instead preferring to focus on its effects. How do you understand hybrid sovereignty, and do you believe your conception centres sovereignty as an important subject for more careful academic consideration across the field of IR?

Sovereignty is a sponge-concept as it stands in for many contested meanings. Some prefer if we abandon its use altogether. I disagree. We cannot wish away sovereignty’s role in structuring international politics. Instead, I present debates about sovereignty as the interplay of two contrasting modes. In Idealized Sovereignty, sovereign authority is represented exclusively as public in “the state” per the doctrine of indivisibility developed by early modern theorists. In Lived Sovereignty, achieving sovereign competence involves divisible practices of state and non-state actors in a variety of public/private social relations. Unlike most standard treatments, I argue that we would do a disservice to sovereignty’s complexity if only one of the two modes prevailed. Instead, sovereignty should be hybridized as both idealized and lived at once. We may idealize sovereignty as indivisibly public, but live sovereignty as divisibly public/private.

Private or non-state influence in sovereign politics often leads to asking whether sovereignty is being “eroded” or the state is in “decline.” In hybrid sovereignty, public/private hybridity is both integral to sovereign power and a challenge to sovereign authority. Treating sovereignty as hybrid then enables IR scholars to embark on more meaningful dialogues about the future of sovereign governance.

You have outlined how hybrid sovereignty existed even before the state system we know today existed. Is the dynamic between private and public actors which creates this hybridity significantly different today than it has been throughout modern history?

While hybrid sovereignty is not new, the political stakes differ throughout history. For instance, in early modern Europe, when I examined how the English East India Company (EIC) became a corporate sovereign, hybrid sovereignty was tolerated for over 125 years until the EIC’s “sovereign awakening” was made unacceptable through public deliberation.

Today, we see similar forms of public/private hybridity as the EIC’s occurring through contracts, institutional linkages, or shadow relations, but the political stakes of hybrid sovereignty have changed. One reason is that Idealized Sovereignty is less accommodating of overlapping and layered sovereign authority. Another reason is that practices of Lived Sovereignty might be more visible, either because of widespread use or due to publicity from information and communication advances, and this increased visibility then becomes politicized, such as the case of war contractors. But these are all historically contingent contests. In another hundred years, we might imagine different political stakes, such that contractors are no longer seen as threatening to Idealized Sovereignty.

Is it useful to think about the power of big private companies like Facebook or Google as manifesting like a government or does the power they have manifest in different ways?

This is the question I take up in my current book project, so I can only offer preliminary thoughts. I became interested in Big Tech when I was writing the conclusion to the first book. I wondered then, “who is the East India Company of today?” Facebook was in the news because of the Cambridge Analytica scandal and Google was the main target of the emerging literature on Surveillance Capitalism.

There are a lot of similarities between Big Tech companies and the EIC: global reach, resource extraction, North versus South dynamics, unwilling governors, and democratic unaccountability, to name a few. But there are also distinctive features of how Big Tech has interacted with states and their users that we have not seen before. The tech companies also do not mimic governments entirely in their governing power. I will present an argument capturing what is the same and what is different in the next book, so stay tuned!

At time of writing, the US has gotten closer to a TikTok ban over concerns about privacy and national security, however many young people on TikTok also see it as a useful platform for advocacy and mobilization. To what extent should attempts to regulate social media be considered as interference with democratic rights to freedom of expression? 

I have spent this year on a fellowship at the Institute for Rebooting Social Media at Harvard, where I have had the chance to learn there are many debates on social media regulation including whether a TikTok ban is valuable. Clearly, many view TikTok and other social media platforms as integral to freedom of expression online. This ban is similar to what Facebook dealt with more than a decade ago when its service was being shut down during the Arab Spring. At that time, liberal democracies saw government social media shutdowns as a problem, but now they are themselves pursuing such policies.

For my research, I am interested in how Big Tech platforms present themselves as defending human rights, including freedom of expression, against governments to their users. However, at the same time, these platforms violate human rights themselves, including freedom of expression through their content moderation systems. This balancing act is one of the differences I alluded to in the previous question of what makes Big Tech different from the East India Company.

States like India, the US, the UK, and many others are considering the use of A.I. to aid in government decision making. How is this use of A.I. likely to change governing authorities’ relationships with their populations?

In “algorithmic governance,” governments defer aspects of decision making to AI systems, sometimes with human supervision. There has already been pushback against this, such as in the Netherlands when AI systems used to determine child welfare payments were found to be discriminatory and in many cases lead to removal of children. In response, an “explainable AI” movement argues that affected populations should (1) have a right to know that AI is making (or helping make) governance decisions and (2) a right to be explained the reasons behind algorithmic inferences. However, there are technical problems with carrying this through, including not knowing what level of explanation would be satisfying to someone denied housing or employment from interacting with AI systems. More generally, some argue that AI systems may be no more biased than humans, who also do not explain their reasoning.

I find that the more interesting political questions are not just about AI bias and discrimination, but about how algorithmic governance creates a new authoritative logic where we accept more AI integration into our daily lives. In this case, it is not just government use of AI but the presence of algorithmically-infused societies that make it difficult to parse out what harms are specific to AI and what are not.

Currently, EU lawmakers are trying to regulate A.I., yet they are also racing to fund A.I. startups within Europe. This reflects a wider trend across the world to gain more control of technologies like A.I. How realistic are the EU and other governing bodies’ attempts to exert more digital sovereignty, and what might be the outcomes of their attempts to do so?

If sovereignty is a sponge-concept, digital sovereignty may be doubly so. These contrasting government efforts actually reflect hybrid sovereignty quite well. In Idealized Sovereignty, states attempt to exert sovereign control over AI companies through regulations such as the E.U.’s AI Act that bans certain uses of AI and creates additional obligations for high-risk AI systems. In the U.S., export controls and the CHIPS Act are an example of this as well since AI advances depend on high-performing computer chips. However, in Lived Sovereignty, states exert sovereign competence through AI companies such as in partnerships for data analytics or storage and incentivizing hardware innovations, all through funding startups or easing regulations.

So, when we ask about the feasibility of achieving “digital sovereignty,” the question supposes a clear meaning of the term that does not exist. Idealized Digital Sovereignty is apparent in the project for a “European sovereign cloud,” but in Lived Digital Sovereignty that vision can only be realized through involvement of American companies. A hybrid sovereignty approach makes sense of seemingly contradictory policy choices in the name of “digital sovereignty.” In short, we should expect to see states both assert control over AI and control through AI.

What is the most important advice you could give to young scholars of International Relations?

Read broadly beyond the IR canon and generate original data. The first will help you in theory-building and the second will help in world-building. Both are important to advance our field.

Further Reading on E-International Relations