PREDICTIVE LAW ENFORCEMENT AND BEHAVIORAL THREAT ASSESSMENT
A neutral guide to behavioral threat assessment, predictive policing algorithms, behavioral-science investigative units, watchlisting, traveler-risk analysis, and the governance of anticipatory state action.
ANALYTICAL & SAFETY BOUNDARIES
Analytical boundary: “Pre-crime” is an analytical label for anticipatory decision architectures, not a single technology or proof that future conduct can be known. Record retrieval is not prediction; identity matching is not prediction of conduct; association is not guilt; a map is not necessarily a forecast; a concern category is not a statistical probability; a score, tier, list, or alert is not lawful grounds for coercive action; formal human presence is not meaningful review without time, evidence, authority, and an effective intervention path. Exact source bytes are preserved, but external claims remain source claims until independently verified.
Simulation safety boundary: Educational and non-operational. No real-person scoring, surveillance deployment, list construction, targeting, coercive workflow, evasion, or system-defeat guidance.
Source basis: 16 exact owner-supplied source instance(s) preserved in the WIP.75 predictive-law-enforcement collection.
ORIENTATION
“Pre-crime” rhetoric can collapse very different tools into one imagined capability. A defensible analysis follows the full decision architecture from data collection to intervention, retention, correction, and redress.
WORKING BRIEF
The relevant chain is collection, identity resolution, retrieval or inference, prioritization, human interpretation, intervention, retention, and review or redress. An algorithm may be absent while the architecture remains strongly anticipatory; a sophisticated model may be present while the output remains low-consequence planning information.
Behavioral threat assessment and management examines a concrete concern through structured professional judgment and continuing case management. Predictive policing algorithms estimate future place or person risk. Behavioral-science investigative units provide research, consultation, interviewing, case formulation, training, and operational support. They overlap institutionally but should not be treated as one predictive machine.
The same score can be low consequence when used for aggregate research and high consequence when it directs stops, home visits, screening, detention, exclusion, or information sharing. Evaluation must therefore include downstream actions, not only model metrics.
COMPLETE DOSSIER
No related public page is required to understand this analysis.
Terms are defined for this site’s evidence method, not as universal legal or clinical definitions.
| Function | Typical output | What it does not establish | Primary governance question |
|---|---|---|---|
| Record retrieval | Matching stored records | Future conduct | Are records accurate, relevant, authorized, and correctable? |
| Identity resolution | Probable entity match | Guilt or threat | How are ambiguity, aliases, and redress handled? |
| Place forecasting | Area/time risk estimate | Who will offend or whether patrol reduces crime | Does the forecast add value beyond transparent hotspot methods? |
| Person scoring or prioritization | Rank, tier, list, or probability-like output | Lawful grounds for intervention | What features, thresholds, consequences, retention, and appeal apply? |
| Behavioral threat assessment | Contextual concern and management plan | Certain prediction of violence | Is the inquiry individualized, multidisciplinary, dynamic, and service-capable? |
| Watchlisting or screening | Match, handling code, screening status, or review request | Charge, warrant, or universal action command | Who nominated, reviewed, acted, retained, corrected, and provided redress? |
Is the system estimating a place, event, identity match, victimization risk, offending risk, vulnerability, intent, or information location?
Pass condition: The target variable and time horizon are explicit.
What operational action follows the score, list, map, flag, or recommendation?
Pass condition: Every downstream consequence and legal authority is identified.
Are outcomes compared with an equivalent untreated group and separated from police attention?
Pass condition: Precision, recall, calibration, false positives, subgroup performance, and causal limits are reported.
Can intervention-generated contacts or arrests become future model inputs?
Pass condition: Generated data is identified, discounted, corrected, or isolated.
Can a person discover, challenge, correct, delete, or appeal consequential data and designations?
Pass condition: Notice, retention, correction, deletion, appeal, and independent review are documented.
RESEARCH EDITION
This page follows the public method for provenance, confidence, source independence, alternative accounts, limitations, review state, and visible correction.
CONTINUE