Police officers standing by a car

The digitalisation of the police is not neutral, but political

Furthermore, the new digital technologies embed different kinds of notions about the future and political choices, says Vasilis Galis from the IT University in Denmark. He has led the large, international research project, CUPP (Critical Understanding of Predictive Policing), which is soon to be concluded. The aim of the CUPP project is to critically engage with the implications of new technologies and advanced data integration and analysis in relation to police work in Denmark, Norway, Sweden, Estonia, Latvia, and the United Kingdom.

Galis says that the choice of whether to outsource the development of digital tools or develop them in-house has been the subject of some serious discussions in terms of what’s best and most optimal and efficient when it comes to law enforcement, as well as in terms of issues concerning privacy and security risks. Ultimately, some Nordic governments have decided to develop digital tools in-house, while others have outsourced their development.

Galis explains that the governments in the countries that have procured new digital technologies have tried to approach the digitalisation process with some kind of neutrality and scientific validity:

“When you start scratching the surface, you realise that specific political doctrines such as neo-liberalism, the claim of efficiency, or the desire to do things rapidly, has caused a lot of trouble within police organisations, and people have reacted negatively to these procedures both when it comes to the procurement process as well as to new working practices.”

The challenges of outsourcing

Another example of the non-neutral and political influence of the digitalisation of the police force is the choice of who is being selected to develop the new software.

“In Denmark and Norway, the police have chosen to work with the American company Palantir, and this choice has indeed been treated by both the local and international press as being politically motivated and not neutral, given the reputation of this large company and its partnerships. Even if you look at the international debate on big-tech, predictive policing, digital policing, and intelligence-led policing, it has not been uncomplicated or neutral,” Galis explains.

He goes on to state that while co-operation between the police departments and Palantir has seemingly been flawless in Denmark, in Norway, there have been significant problems. Here, the police organisation chose not to purchase the complete system from Palantir, but instead decided to develop parts of the system in-house which, among other things, caused a lot of unnecessary expense.

And then you have Sweden, says Galis:

“In Sweden the police chose to undertake part of the development in-house. At the same time, the public debate surrounding and insight into the digitalisation of the police has been quite limited for a Nordic country that is so vocal about its openness and public transparency. However, more and more journalists have become very interested in the process of the digitalisation of the police, with a keen emphasis on the current so-called gang crime crisis. This has given rise to debate concerning the safety of the general public, which is connected to the potential to solve this crisis digitally. However, in this, you can see how specific events are instrumentalised in order to accelerate the transformation of the technological arsenal of the state and law enforcement, regardless of the relevance of those tools for actually solving the problem.”

The debate on this issue is two-dimensional, according to Galis:

“If you look at the big picture, although general crime rates in Scandinavia are going down, in Sweden there’s a real emphasis within politics and the media on a specific type of crime that’s being given a lot of attention, and this has created demand for more policing, more surveillance, and more digital solutions.”

Galis states that the demands, to some extent, have been railroaded through, as politicians and the media require fast and spectacular solutions. However, these solutions and the pace with which they have entered the public debate and administration, have been at the expense of issues concerning privacy, surveillance, human rights, and the way states are penetrating our digital communications. And this is very common in all the countries that CUPP has investigated, he says.

Furthermore, Galis states that specific narratives have been used to intensify the outsourcing and procurement of new digital tools:

“As an example, the narrative of the ‘War on Crime’ has intensified, and the discussions connected to ‘The War on Something’ have become very militarised. Although this has pacified the Scandinavian societies considerably, resulting in greater acceptance of the new technologies and lack of public procurement procedures in the name of ‘The War on Something’, it has nevertheless led to a situation where the digitalisation process has become a confined and non-transparent process.”

Vasilis Galis, Associate Professor at the IT University of Copenhagen and leader of the research project "Critical Understanding of Predictive Policing (CUPP)".

A new set of semantics and categorisation of the population

Galis describes the digitalisation of the police’s work as very transformative:

“The work of the police is about to change. Although police officers have always had to collect data after a crime has been committed, the new digital tools now allow them to also use preventative methods. This means that the police do not necessarily need to react to a crime, but instead they engage a new breed of analysts sitting at computers, using data to produce reports and forecasts to predict future events.”

Therefore, existing methods and practices have collided with or are complemented by the new paradigms:

“Of course, although the police have always been involved in surveillance and tried to produce prognoses, the new digital tools have accelerated the pace of analysis, and they have actually provided the police with a degree of efficiency, as police officers now spend much less time on manually trawling through databases in ongoing investigations. Nevertheless, the structure of the computer software is not based on an objective or scientific setup, rather it’s primarily based on semantics and choices made by private stakeholders and representatives of the police organisation.”

During the study of one of these companies (the company mentioned earlier in this article, Palantir) Galis and his colleagues were quite surprised when they stumbled upon a very controversial term, “ontology”:

“Palantir said ‘we’re developing an ontology’. In philosophy, ontology has a very specific meaning of how we perceive the world. Social scientists look at ontology as a part of creating a reality. For us, listening to data engineers saying that they were creating an ontology, immediately raised some red flags,” Galis says, before he continues:

“Studying the concept, use, and implementation of ontology even further, one quickly realises that software companies in co-operation with state and police authorities have created a whole world of semantics, that is to say a world of interpretation with classifications, categorisations, and options embedded in the software systems. And it’s these systems that are prompting the police to perform in a certain way.”

As an example, Galis uses a police patrol that has been called to an area where a crime is suspected. When the police officers go to their mobile phones or tablets to register something, they’re exposed to certain options such as burglary, attempted murder, or something similar.

“When you create a model of options to choose from, you describe the world in a certain way, but also exclude other semantics and options. Although the option may be very broad, it still excludes other understandings or categorisations of crime. This kind of creating of a reality through the semantics embedded in the systems, that is to say the data science ontology that law enforcement is exposed to, creates certain worlds with a specific set of classifications.”

According to Galis, the CUPP research indicates that because of these classifications, certain crimes are prioritised over others, such as gang crime over domestic violence. Consequently, digital tools are not only technologies that describe the world, but they also influence the world. Galis concludes by stating:

“Our ambition for the impact of our research project is to shake up the societies and make them aware that there are other alternatives to the ‘new’ way of policing. Because what governments and police authorities have put forward as a very scientific and neutral procedure is in fact not neutral. However, there are alternatives, and societies ought to explore such alternatives that will hopefully provide more transparency and respect basic human rights.”

Read more about the CUPP project in our newly published Fast Track to Vision 2030 (pages 28-32)


Bodil Aurstad. Photo: NordForsk

Bodil Aurstad

Special Adviser