The Research Problem
Crime scene investigation is a critical starting point in any police inquiry, where forensic practitioners must make carefully considered decisions based on what they observe. These initial decisions govern investigative directions. However, practices vary widely across regions due to, for instance, differences in cultural and legal frameworks, access to resources and training, and protocols. This is especially evident in cases of violence against women, where investigative priorities, resource allocation, and inherent investigative biases may significantly undermine the fairness of justice outcomes. This project explores whether Generative Artificial Intelligence (GenAI) can help improve fairness and consistency in how such crime scenes are investigated.
By analysing how decisions vary across regions, languages, and cultural contexts, this project seeks to identify patterns of bias or inconsistency in both human and GenAI-based scene reconstructions. This will help us understand whether GenAI can support more consistent and equitable practices and what ethical or practical challenges might arise from its integration into forensic workflows.
Research Design
Research Question: How do GenAI outputs compare to decisions from forensic practitioners, educators, and students from different countries when reconstructing identical simulated crime scenes of crime against women?
Methodology: WICTORIA foundational step is the creation of simulated crime scenes representing gender-based violence offences against women. Following their immersive exploration of these simulated crime scenes, participants will complete a survey to capture their decision-making processes. Participants will also be asked to compare their own selection of traces and reconstruction hypotheses with those proposed in the AI-generated compendium. This will enable capturing their reasoning, decision-making strategies, and reflections on investigative priorities. This behavioural data will provide rich insights into how professional background, experience, and culture shape forensic reconstruction of gender-based violence offences.
Partnership details:The project brings together an interdisciplinary and international group of researchers with expertise spanning forensic science, statistics, cognitive psychology, computer science, gender-based violence, discourse analysis, and the law.
Project Objectives
WICTORIA brings together forensic science, gender-based violence studies, and artificial intelligence. It aims to understand how individuals from different cultural, linguistic, and professional backgrounds detect and interpret traces when processing crime scenes and reconstructing gender-based violence offences, and to explore whether generative AI (GenAI) can enhance this process by promoting fairness, consistency, and inclusivity.
WUN’s distinctive strength lies in its access to a culturally and geographically diverse network of expert researchers, educators, and practitioners. This project draws on that diversity to examine how forensic scene reconstruction and decision-making processes vary across different regions, languages, professional cultures, and levels of expertise.