Future crime spots or potential perpetrators can be predicted with forecasts that are based on algorithms. Postdoctoral Researcher Carlo Gatti from the University of Turku in Finland studied the legitimacy and use of predictive policing in his home country Italy.

Predictive policing refers to the use of algorithmic predictions to forecast police targets, such as future crime spots or potential perpetrators. This phenomenon has attracted increasing attention in recent years.

Postdoctoral Researcher Carlo Gatti from the University of Turku studied the topic in his doctoral dissertation. He examined the policies and rationales for predictive policing from a multidisciplinary perspective.

“Predictive policing is justified on both criminological grounds and with more surface-level explanations," says Gatti. “There are different kinds of approaches for predictive policing.”

“In person-based approaches, algorithms estimate individuals' risk of committing future crimes based on selected characteristics. Whereas in place-based approaches, the system analyses crime history and environmental factors to predict where crimes are likely to occur.”

Some systems are developed by police authorities, such as CAS in the Netherlands. Others are commercial tools provided by global technology companies — such as Palantir — and tailored to different national contexts.

“Although person-based approaches are often considered more intrusive, my research shows that place-based approaches can also be built on sensitive data and produce similar effects. It raises critical questions about the effectiveness of public control.”

Soft privatisation can be a problem for democracy

“There are legal gaps and regulatory shortcomings in predictive policing," says Gatti.

In his research, Gatti analyses the transparency of public and private actors, and identifies legal shortcomings from the perspective of actual systems.

"My research shows that regulatory countermeasures fall short of the goals they claim to pursue.”

The study also examines the relationship between private and public power. According to Gatti, soft privatisation undermines the accountability of public authorities that is part of democracy.

Gatti studied predictive policing, particularly in the context of his home country Italy. He analysed the properties of two closed algorithms used in Italy.

“There has been little previous research on this topic in Italy and no systematic data from developers or users of predictive policing."

“Critical evaluation of the impact of such technologies is important before their possible implementation,” Gatti emphasises.