Predictive Policing & Social Scoring is the use of algorithms to forecast crime and rate people by risk or trustworthiness. It overlaps social credit AI and the surveillance state.
What is documented
Predictive-policing tools and risk-assessment systems such as COMPAS are documented and heavily studied — including their well-known tendency to encode the bias in the data they learn from. Feedback loops, where policing the predicted areas generates the data that confirms the prediction, are a recognized problem.
How the map holds it
The map takes the documented harms — biased data, opaque scoring, self-fulfilling loops — seriously on their own terms, while resisting the inflation of fragmented tools into one finished global scoreboard. Accuracy about how far these systems actually reach makes the critique stronger, not weaker.
Whether you see it as flawed tooling or the seed of a scoring state, the map’s standard is the same: no consequential score without transparency, contestability, and an audit of its bias.