Prediktive politi- og etterforskningsverktøy: En komparativ analyse av dansk og norsk regulering
Translated title
Predictive Policing- and Investigative Tools: A Comparative Analysis of Danish and Norwegian Regulation
Author
Term
3. term
Education
Publication year
2025
Submitted on
2025-12-01
Pages
50
Abstract
Artificial intelligence is gaining importance in police and investigative work. Predictive analysis tools enable the police to process and analyse large amounts of data. Through algorithmic assessments, some systems can identify patterns and potential risk areas for crime, increasing the efficiency of police work. On the other hand, it is questionable whether fundamental principles of legal protection can be safeguarded when such analytical tools are used. This thesis examines the extent to which predictive policing and investigative tools can be used within Danish and Norwegian law without violating the right to adversarial proceedings, verifiability, and the presumption of innocence. As Article 6 of the European Convention on Human Rights provides important human rights standards for these principles, it will also be included in the analysis. To answer the research question, the thesis analyses the applicable law in Denmark and Norway, followed by an assessment of how each jurisdiction addresses the challenges of relying on AI-based assessments in criminal evidence and decision-making. The analysis shows that both Denmark and Norway, in principle, allows for the use of predictive tools, but they take notably different approaches. Denmark has adopted a broad, function-based legal framework that gives the police wide access to implement such analytical tools as part of ordinary investigations. In contrast, Norway has adopted a far more restrictive approach, limiting the use of such systems to the Norwegian Police Security Service. A common feature in both countries is the absence of detailed procedural rules that describe how algorithmic assessments should be controlled and evaluated. When specific rules for this technology do not exist, the general legal safeguards must protect the individual. Rights such as access to information, the opportunity to challenge evidence, judicial control over the presentation of evidence, and the requirements of necessity and proportionality in data collection are essential safeguards. Together with the presumption of innocence, these mechanisms function as important guarantees of legal protection when such tools are used. These principles set boundaries for how algorithmic analyses can be used, presupposing the technology is designed and practiced in a way that makes transparency and verifiability possible. The thesis concludes that predictive policing and investigative tools may be used within both legal systems, but only under clearly defined conditions. Although the technology may be useful for increasing efficiency in investigations and improving police work, it must always be regarded as a support tool and not as an independent basis for decisions. Human judgement and control are essential to prevent algorithmic assessments from gaining too much influence in the criminal process. To maintain legal safeguards, the use of artificial intelligence must be transparent and compatible with the fundamental procedural principles that form the basis of the rule of law.
Documents