It’s time to regulate police use of crime prediction algorithms
Additionally, the algorithms generate feedback loops, so that when an error results in a penalty (for example, if an application incorrectly registers a person in a database as a suspect), the error is copied to other other sets of data until it becomes impossible to identify. where the first mistake was made.
In recent years, police and probation services have increasingly used automated services to help identify suspects or predict recidivism rates. In a report published last month, the UK Parliament’s House of Lords Justice and Home Affairs Committee gives examples: Avon and Somerset Police bought a tool called Qlik Sense to predict crime trends; the Home Office uses an algorithm to review applications for marriage licenses to predict which marriages are “sham” and thus justify denial of visas; and Durham Constabulary uses a risk assessment tool, which uses machine learning to predict who is likely to commit crimes and should therefore be watched more closely.
The commission speaks of the “benefits” of these regimes in terms of increased efficiency. He warns, however, that technologies of punishment and control are developing faster than legal resources to prevent injustice. For example, the committee says it’s unusual for a criminal defendant to be told if their investigation is based on a recommendation from a computer application. He also speaks of a “vicious cycle,” in which an app preemptively identifies groups of people living in particular areas as being at risk of becoming criminals, causing police to closely monitor their lives, with this surveillance generating in its turn turn of the data that warrants further investigation.
The authors note, but do not properly explain, that several of the algorithms recently purchased by police forces mimic others that have been controversial. A highly published application, widely used in the United States (and in Kent between 2013 and 2018), is PredPol, which uses historical crime data to predict, hour by hour, where a crime might be committed. Police forces are issued maps with boxes marked in red to indicate where crime should be expected. Sites of previous burglaries can be patrolled and possible crimes prevented until the incidence of that offense – and crime in general – has decreased.
We can see the legacy of PredPol in some of the newly acquired algorithms. QlikSense, used in Avon and Somersetmimics its focus on the local geography of crime and its changing local trends.
There are, however, good reasons why PredPol has become controversial. First, its proponents have insisted that predictive policing will make ordinary citizens safer. This claim is, in turn, based on the theory of “zero tolerance”: that if only you could eradicate petty crime and antisocial behavior, serious crime would also decrease, and both could be achieved without social cost. . “Criminal offences”, according to the formation of Predpol materialsare “considered the gateway to more serious crimes”.
In his international bestseller “Humankind”, Rutger Bregman showed how poorly thought out the studies on which the zero tolerance policy was based. It began with a series of experiments in which psychologists deliberately created conditions designed to maximize the incidence of crime, and then recorded (unsurprisingly) that it increased. Meta-analyses of zero-tolerance policing show that there is in fact no correlation between its aggressive approach to petty crime and reduced violence. Increasingly arbitrary and authoritarian policing – inevitable with a zero-tolerance approach – produces not fewer offenders but more disgruntled people.
The problem is that the most minor offences, such as petty theft or drug possession, are those whose application is most distorted by demographics. A middle-class teenager is more likely to have their own bedroom, perhaps some distance from their parents’ bedroom, where they can smoke cannabis or drink alcohol before they are underage. A working-class teenager shares more often with siblings and is therefore more likely to smoke or drink on the street. The middle-class criminal is invisible to app-based policing; the worker criminal is seen and punished. Being black and working class puts you even more in the eyes of the police.