Can there be too much automation?
This is one the questions the Critical Data Network discussed in relation to the use of digital data for police and intelligence work.
From the 3rd CDN workshop at the Faculty of Law. Photo: UiO.
Police work has always been data-intensive: collecting, organizing and analyzing information is highly characteristic of police procedures. With the rise of dataveillance and predictive policing technologies data analysis has begun to re-structure police work. Even though the police have access to more information today, a key task remains to make sense of this information.
The Oslo Police, the Financial Intelligence Unit and digital data
Carl Erik Bergwitz-Larsen, an intelligence officer at the Oslo police, discussed the promises and pitfalls of digitization and automation in intelligence work with members of the Critical Data Network. Today, the Oslo Police and the Financial Intelligence Unit use software systems mainly for decision support. Such systems are designed to organize information rather than automatizing decision-making. And yet, the organization of information is a core part of the interpretation of information as well as the concrete decisions that officers take. Or to put it in expert Heidi Mork Lomell’s words: “it influences how police officers see things”.
Kaufmann, lead of the network, explains: “the challenge is here not so much that police officers feel replaced by a machine, but rather that digital modes of sorting and categorizing information steers the police gaze and police practices. Building such systems is then a critical moment of police work and needs to be executed with extreme care.”
“In that sense”, Kaufmann adds, “the main task is probably less to identify a perfect balance between automated and non-automated processes in the police. Instead, we need to capture the ways in which digital, analogue and human forms of data processing will always influence each other and create concrete effects in police work and society.”