# Privacy and Civil Liberties Predictive intelligence is valuable precisely where evidence is incomplete. That is also where false positives, identity errors, base-rate problems, hidden assumptions, and automation bias can do the most harm. The safeguards route includes [[wiki/Intelligence Oversight|Intelligence Oversight]], [[wiki/Privacy and Civil Liberties|Privacy and Civil Liberties]], [[wiki/U.S. Person Information|U.S. Person Information]], [[wiki/Minimization|Minimization]], [[wiki/Purpose Limitation|Purpose Limitation]], [[wiki/Data Retention|Data Retention]], [[wiki/Authorized Use|Authorized Use]], [[wiki/Query-Based Access|Query-Based Access]], [[wiki/Auditability|Auditability]], [[wiki/Watchlist Redress|Watchlist Redress]], [[wiki/Misidentification|Misidentification]], [[wiki/Due Process|Due Process]], [[wiki/Human-in-the-Loop|Human-in-the-Loop]], [[wiki/Explainable AI|Explainable AI]], and [[wiki/Prediction versus Determination|Prediction versus Determination]]. Public-health modeling sharpens the same distinction. [[wiki/Human as Organism and Person|A human can be modeled as organism, node, receptor, carrier, contact, or cohort member while remaining a legal Person]]. Observability attaches to the phenomenon and representation; individualized state action encounters authority, evidence, process, and constitutional constraint. Civic literacy must resist two errors at once: treating every security capability as illegitimate surveillance, and treating a security purpose as automatic authorization for every use. [[wiki/Probabilistic Precrime|Probabilistic precrime]] sharpens the boundary. Retrospective policing asks what may follow from an act that occurred; prediction-led policing asks what may be observed, inferred, prioritized, investigated, or interrupted because a harmful future has become more probable. Person-specific probabilities are neither acts nor intentions nor proof of guilt. Provenance, model validation, bias testing, contestability, and intervention-specific legal authority must remain visible at the decision point. The harm-accounting route now has two layers. [[wiki/Security Harms Topology|Security Harms Topology]] is the master map of root-state, propagation, encounter, intervention, trajectory, epistemic, relational, institutional, and residual harms. [[wiki/The Second Error Function|The Second Error Function]] is the ledger that measures those harms and returns them to the protective system as negative feedback. [[wiki/Watchlist Misidentification|Watchlist Misidentification]] supplies the clearest distributed case: [[wiki/Identity Collision|identity collision]] and [[wiki/Erroneous Watchlist Inclusion|erroneous inclusion]] can become a [[wiki/Distributed Erroneous State|distributed erroneous state]], accumulate [[wiki/Cumulative Downstream Burden|downstream burden]], and reorganize a [[wiki/Life-Trajectory Harm|life trajectory]] before correction reaches every system. [[wiki/Redress Opacity|Redress Opacity]], [[wiki/Legibility-Gated Redress|Legibility-Gated Redress]], [[wiki/Surveillance Purpose Migration|Surveillance Purpose Migration]], and [[wiki/Surveillance Trust Externality|Surveillance Trust Externality]] supply additional documented terms, and [[wiki/Operational Activity and the Common Rule|Operational Activity and the Common Rule]] locates the governance seam between operational surveillance and human-subjects research. The full analysis is [[research/Harms Incurred While Bringing Preventive Systems Online|Harms Incurred While Bringing Preventive Systems Online]]. ## Relationships [[wiki/Counterterrorism Collection - Public Health and Violence Prevention|Public Health and Violence Prevention]] · [[wiki/Counterterrorism Collection - Fusion and Surveillance|Fusion and Surveillance]] · [[wiki/Counterterrorism Collection - Predictive Policing and Probabilistic Precrime|Predictive Policing and Probabilistic Precrime]] · [[wiki/Observational Modeling versus Individualized State Action|Observation versus intervention]] · [[wiki/Security Harms Topology|Security Harms Topology]] · [[wiki/The Second Error Function|The Second Error Function]] · [[wiki/Watchlist Misidentification|Watchlist Misidentification]] · [[research/Harms Incurred While Bringing Preventive Systems Online|Harms Incurred While Bringing Preventive Systems Online]]