Algorithmic governance of ‘terrorism’ and ‘violent extremism’ online

Abstract Global security risks are increasingly countered through complex data infrastructures involving forms of algorithmic governance, automated decision-making and data-driven, public-private co-regulation. This paper explores the challenges these security infrastructures pose by following a key regulatory device in the global governance of ‘terrorist and violent extremist content’ (or TVEC) online: the hash-sharing database of the Global Internet Forum to Counter Terrorism (GIFCT). By drawing on interviews with platform counterterrorism experts and closely examining the materiality of hash-sharing practice, this infrastructure is analysed as an emergent sociotechnical assemblage for countering ‘TVEC’ that is generating far-reaching effects. The paper shows how it is reconfiguring legal and security governance practices, forging novel forms of power and material agency, and generating distinctive algorithmic ordering techniques and ‘infra-legal’ dynamics that conventional law and technology research has not yet adequately explored. Grappling with these effects requires foregrounding material processes, decentring law and being empirically attentive to how law and data infrastructure are enmeshed and co-evolve in practice. The paper suggests new possibilities for the sociolegal study of global algorithmic infrastructures in action.