# Prediction Is Not Jurisdiction **Domain:** Rights / Prediction / Due Process **Doc Type:** Canonical Concept Node **Maturity:** Evolving **Related:** [[wiki/Behavioral Prediction|Behavioral Prediction]], [[wiki/Predictive Modeling|Predictive Modeling]], [[wiki/Algorithmic Governance|Algorithmic Governance]], [[wiki/Appeals Mechanisms|Appeals Mechanisms]], [[wiki/Rights|Rights]] --- ## Definition **Prediction Is Not Jurisdiction** is the principle that increased confidence about a future act may inform investigation or precaution without automatically authorizing coercion, punishment or irreversible deprivation. Forecast accuracy changes the epistemic situation; it does not erase the legal and moral distinction between possibility, preparation and completed conduct. ## Constitutional Function The principle is not a prohibition on acting before harm. Imminent threats can justify proportionate intervention. What it prohibits is conversion of a probabilistic model into an unappealable verdict. A defensible system must expose the prediction's evidence, uncertainty, time horizon, alternative explanations and the authority under which any response occurs. [[wiki/Person of Interest S3E20 — Death Benefit|Person of Interest S3E20 — Death Benefit]] supplies the direct case: a forecast of catastrophic future action becomes a proposed death sentence. [[wiki/Person of Interest S4E5 — Prophets|Person of Interest S4E5 — Prophets]] supplies the recursive case: an administrator capable of altering reality can manufacture the outcome that validates its forecast. [[wiki/Westworld S3E7 — The Outlier Problem|Westworld S3E7 — The Outlier Problem]] supplies the population case: unpredictability becomes grounds for removal. ## Real-World Interface The EU AI Act prohibits specified forms of individual criminal-risk assessment or prediction, reflecting the broader rule that people should not be judged solely through predicted future behavior. This is a bounded legal analogue, not a universal ban on predictive tools. ## Forecasts as Causal Interventions [[articles/Peak Person and the Predicaments of Prediction|Peak Person and the Predicaments of Prediction]] sharpens the principle: once an institution reallocates opportunity because of a forecast, the forecast becomes [[wiki/Institutional Retrocausality|Institutional Retrocausality]]. [[wiki/Forecast-Induced Dissipation|Forecast-Induced Dissipation]] occurs when withdrawal produces the decline that the model then claims to have predicted. The right therefore includes [[wiki/Counterfactual Opportunity|Counterfactual Opportunity]]—some protected access to the job, treatment, education, relationship, repair or environment capable of falsifying the model. This does not prohibit precaution or final decisions. It prohibits making a model unfalsifiable by closing every path the model did not expect. [[articles/Ambiguity Will Destroy Man and Machine|Ambiguity Will Destroy Man and Machine]] adds [[wiki/Epistemic Status Separation|Epistemic Status Separation]]: observation, suspicion, prediction, finding, sentence and permanent identity must remain different administrative objects. ## Key Insight **A forecast can justify attention without becoming a sentence.** ## See Also [[wiki/World Modeling|World Modeling]], [[wiki/Operator Sovereignty|Operator Sovereignty]], [[wiki/Score Separability|Score Separability]], [[wiki/Person of Interest S2E16 — Relevance|Person of Interest S2E16 — Relevance]], [[wiki/Trajectory Governance|Trajectory Governance]], [[articles/Peak Person and the Predicaments of Prediction|Peak Person and the Predicaments of Prediction]] ## Sources / Provenance - [European Commission: AI Act](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai). - Fictional instrumentation: [[wiki/Person of Interest Episodes|Person of Interest Episodes]] and [[collections/Westworld Exploration|Westworld episode exploration]].