Looking Through the Looking‐Glass: Digital Authoritarianism in (Regulated) Practice

Debates on ‘digital authoritarianism’ have tended to focus on authoritarian regimes and their deliberate use of digital technologies for repressive ends. This article argues that such a focus obscures as much as it reveals. Drawing on practice theory and material semiotics, it proposes a reconceptualization of ‘digital authoritarian practices’ that treats digital technologies not as neutral instruments in the hands of human actors but as constitutively entangled in the socio‐material relations through which accountability is sabotaged, whether intentionally or not. To explore what this reconceptualization makes visible, the article examines the European Union's attempt to regulate predictive policing under the AI Act (Regulation EU 2024/1689). This case is illustrative because key regulatory carve‐outs in the Act are not simply loopholes to be closed. They reflect a deeper assumption that digital technologies are mere tools that human actors deploy, rather than active participants in complex assemblages. The article suggests that this assumption renders democratic regulatory frameworks structurally ill‐equipped to address the authoritarian effects they ostensibly seek to prevent. As a concluding invitation, the article puts forward ‘cyborg regimes’ as a conceptual pivot to recast our political imagination and intellectual curiosity.