In the second DSI Insights of the year, Sarah Summers, Director of the Digital Society Initiative (DSI), explains why predictive policing with artificial intelligence (AI) is a challenge for the justice system.
AI-supported systems analyze large amounts of data to predict potential crime scenes or perpetrators. The aim is to prevent crimes before they happen. A distinction is made between focusing on people (e.g. by analysing social networks or previous arrests) and places (e.g. by identifying crime hotspots).
Earlier systems were based on criminological theories and were comprehensible. Newer AI models with machine learning offer greater accuracy, but are often opaque in their decision-making. The use of such AI systems raises questions about discrimination, data protection and legal accountability. There is a risk that existing prejudices will be reinforced by the algorithms.
The author emphasizes the need for a clear legal framework and ethical guidelines to ensure the responsible use of AI in police work.
DSI Insights:
In the “DSI Insights” column series, the DSI demonstrates its expertise and interdisciplinarity on digital transformation topics. The columns by authors from the DSI network appear in the magazine “Inside IT” and on the communication channels of the University of Zurich. Here you can find all DSI Insights since 2016.