Policing efforts to thwart urban crime often rely on detailed reports of criminal infractions. However, crime rates do not document the distribution of crime in isolation, but rather its complex relationship with policing and society. Several results attempting to predict future crime now exist, with varying degrees of predictive efficacy. However, the very idea of predictive policing has stirred controversy, with the algorithms being largely black boxes producing little to no insight into the social system of crime, and its rules of organization. The issue of how enforcement interacts with, modulates, and reinforces crime has been rarely addressed in the context of precise event predictions. In this study, we demonstrate that predictive tools are not purely an instrument enhacing state power, but may be effectively used to seek out systemic biases in urban enforcement. We introduce a novel stochastic inference algorithm as a new forecasting approach that learns spatio-temporal dependencies from individual event reports with demonstrated performance far surpassing past results (e.g., average AUC of ~90% in the City of Chicago for property and violent crimes predicted a week in advance within spatial tiles ~1000ft across). These precise predictions enable equally precise evaluation of inequities in law enforcement, discovering that response to increased crime rates is biased by the socio-economic status of neighborhoods, draining policy resources to wealthy areas with disproportionately negative impacts for the inner city, as demonstrated in Chicago and six other major U.S. metropolitan areas. While the emergence of powerful predictive tools raise concerns regarding the unprecedented power they place in the hands of over-zealous states in the name of civilian protection, our approach demonstrates how sophisticated algorithms enable us to audit enforcement biases, and hold states accountable in ways previously inconceivable.