BiasMedium
Predictive Policing Biases of PredPol
- Occurred
- Nov 18, 2015
- Published
- Dec 2, 2021
- Source
- AI Incident Database
Summary
Predictive policing algorithms meant to aid law enforcement by predicting future crime show signs of biased output.
Developers
- PredPol
Deployers
- PredPol
- Oakland Police Department
Harmed parties
- Oakland Residents
Classification
Category confidence: 95%
Severity confidence: 90%
Rationale: Incident describes biased output in predictive policing algorithms, fitting BIAS category with medium severity due to notable controversy.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8