Edinburgh Napier University

  This paper conceptualises an emerging model of algorithmic policing; ‘influence policing’. This harnesses the affordances of Internet platforms to conduct domestic digital influence campaigns for crime prevention. These campaigns use sophisticated targeted messaging to directly ‘nudge’ behaviour and shape the culture of specific groups. By targeting people using micro-level behavioural, personal-interest, and location-based data, influence campaigns aim to employ insights from behavioural psychology to prevent crime at a distance. We theorise this with an analysis of a dataset of more than 12,000 adverts and in-depth fieldwork with a dedicated police strategic communications team. Influence policing provides law enforcement with new capacities to craft and manicure hidden digital encounters with targeted publics, raising questions its democratic character and police accountability.