predictive policing
In plain wordsUsing computer programs to forecast where crimes are likely to happen, or who is likely to be involved, so that police can decide where to patrol or whom to watch.
The use of statistical and machine-learning models on historical crime, arrest and other data to forecast where and when crime is likely (place-based) or which individuals are likely to offend or be victimized (person-based), in order to allocate police attention. Place-based systems such as PredPol, later renamed Geolitica, adapted earthquake-aftershock models to crime reports; person-based systems such as Chicago's Strategic Subject List scored individuals. Proponents say forecasts help target scarce patrols where crime concentrates; critics argue that because the data record policing rather than crime itself, models send officers back to already heavily policed neighborhoods, which then generate more recorded incidents in a self-reinforcing feedback loop, and that evidence of crime reduction is weak. Several US cities ended such programs from 2019 onward, and the EU AI ActEU AI Act, prop. n.: The European Union's law on artificial intelligence, the first broad AI law anywhere: it bans some uses, sets strict checks for risky ones and puts duties on the makers of the biggest AI models. prohibits assessing an individual's risk of offending based solely on profiling or personality traits.
The city dropped its predictive policing contract after an evaluation found no measurable effect on burglaries.
Origin
The phrase spread in the late 2000s, building on 1990s crime mapping and hot-spot policing: Los Angeles police chief William Bratton promoted forecasting-led policing, and by 2009 the US National Institute of Justice was convening symposia on it. Its first use is not established. PredPol, spun out of UCLA research by Jeffrey Brantingham and George Mohler, launched commercially in 2012.
Usage note
Kristian Lum and William Isaac (2016) simulated a drug-crime model trained on Oakland arrest data and showed it would concentrate patrols in Black and low-income neighborhoods, although public-health surveys suggested drug use was spread far more evenly.
Related
- algorithmic biasalgorithmic bias, n.: Unfairness in software that judges or serves people: it consistently produces worse results for some groups, such as women, older people or ethnic minorities, often because it learned from skewed or incomplete records.
- COMPASCOMPAS, prop. n.: Commercial software used in parts of the US justice system to rate how likely someone is to commit another crime; a 2016 news investigation asking whether it was unfair to Black defendants made it famous.
- AI surveillanceAI surveillance, n.: Using AI to watch, identify, track or profile people on a scale no team of humans could manage, for example by scanning camera feeds or analyzing phone and online data.
- facial recognitionfacial recognition, n.: Technology that identifies or checks who someone is by comparing a picture of their face with stored photos, used by phones, passport gates and police databases.
- prohibited AI practicesprohibited AI practices, n.: Uses of AI that the EU bans outright as too harmful — such as manipulating people, social scoring, and most live facial recognition by police in public places.
- algorithmic accountabilityalgorithmic accountability, n.: The idea that the people and organizations behind automated decision systems must be able to explain and justify those systems' effects, answer for the harm they cause, and fix it.
Sources
- Lum, K. & Isaac, W. (2016). To predict and serve? Significance 13(5), 14–19.
- Ensign, D., Friedler, S. A., Neville, S., Scheidegger, C. & Venkatasubramanian, S. (2018). Runaway Feedback Loops in Predictive Policing. Proceedings of Machine Learning Research 81 (FAT* 2018).
Mentioned in
facial recognitionfacial recognition, n.: Technology that identifies or checks who someone is by comparing a picture of their face with stored photos, used by phones, passport gates and police databases., AI surveillanceAI surveillance, n.: Using AI to watch, identify, track or profile people on a scale no team of humans could manage, for example by scanning camera feeds or analyzing phone and online data., COMPASCOMPAS, prop. n.: Commercial software used in parts of the US justice system to rate how likely someone is to commit another crime; a 2016 news investigation asking whether it was unfair to Black defendants made it famous., feedback loopfeedback loop, n.: A cycle in which what a system produces comes back to shape its own future input, like a microphone howling near its loudspeaker. In AI, a model's predictions can shape the very data it later learns from., predictive AIpredictive AI, n.: AI used to forecast outcomes about people or events, such as who will repay a loan, which patient may get sicker or which machine will fail, in order to guide decisions.