How do predictive monitoring technologies affect the likelihood and form of protest? Classical repression theories analyze state coercion as reactive responses to observable collective action. This framework fails when algorithmic surveillance enables states to intervene before mobilization occurs, inverting the temporal sequence that defines conventional repression. I develop a theory of anticipatory repression wherein states deploy predictive analytics to identify and preempt potential threats rather than responding to manifest challenges. Anticipatory repression operates through three interconnected mechanisms. Perception renders potential dissent informationally visible by capturing behavioral signals through surveillance networks and data scraping. Prediction transforms these signals into probabilistic risk assessments through algorithmic analysis. Preemption operationalizes prediction through early interventions designed to disrupt mobilization before protests become observable in empirical records. These mechanisms generate three testable propositions: predictive surveillance infrastructure deployment (1) reduces protest incidence, with effects strongest for spontaneous, low-threat mobilization; (2) alters protest form by increasing organizational formality and reducing public visibility; and (3) operates similarly across regime types while varying in implementation intensity according to institutional constraints. I employ a nested mixed-methods design combining a quantitative analysis of protest event data with qualitative case studies examining surveillance deployment in China, Nigeria, and the United States. Preliminary analysis suggests that predictive monitoring correlates with reduced mobilization and systematic changes in organizational tactics across different political systems, with significant implications for democratic contestation and civil liberties in an era of ubiquitous surveillance.