Asmolov, G. (2026). The rise of AI sovereignty: Authoritarian technological imaginaries as a form of reflexive control. Big Data & Society, 13(2), 20539517261426
Much of the current debate about the contribution of artificial intelligence to the rise of digital authoritarianism focuses on its role in disinformation, propaganda, and surveillance. But authoritarian influence may also operate at a deeper level, by shaping how AI itself is governed. This question motivated my recent article in Big Data & Society.
The purpose of the article is to explore how authoritarian ideas travel across borders and influence global AI governance. Rather than focusing on the export of technologies and AI-enabled practices, it examines mechanisms of authoritarian diffusion through informational influence, showing how narratives and ideas shape the forms AI governance takes.
The analysis focuses on the rise of “AI sovereignty” as a central concept in global debates. Drawing on the notion of sociotechnical imaginaries and the concept of reflexive control from Soviet strategic thought, the article shows how political actors can influence others by shaping perceptions of technological futures.
Using Russia as a case study, the article analyses authoritarian imaginaries that inform approaches to AI governance. It identifies three dominant narratives: AI as a tool of global power and domination; “Western AI” as a cultural threat; and AI as a driver of state efficiency. It then examines how these narratives may extend beyond Russia, influencing global debates and contributing to the spread of sovereignty-centred approaches to AI governance.
The main argument is that these imaginaries influence the cognitive environment of global policymaking. As they circulate, they promote sovereignty-centred and security-driven approaches to AI governance, even beyond authoritarian contexts.
Recognising this diffusion mechanism highlights the need for greater sensitivity to the potential impact of authoritarian imaginaries on global models of AI governance. Developing such awareness is essential for mitigating the subtle spread of authoritarian logics through policy debates and regulatory frameworks. The article contributes to this effort by advancing a theoretical understanding of how diffusion operates through informational influence, and by showing specifically how it can shape the governance of AI across different political contexts.
