Cross-Sector Collaborations : Challenges and the Role of AI

In recent years, Swedish society has seen a rise in violence, with youth playing a central role in this escalation. In response, the Swedish government, in collaboration with civil society organizations, has launched cross-sector initiatives to strengthen preventive efforts. This thesis is part of the IBob-project, an extension of the nationally coordinated BoB-initiative, which aims to prevent young people from entering a criminal lifestyle by improving collaboration between major governmental actors such as the Police, Social Services, and the National Agency for Education. The IBob-project expands this work by exploring how artificial intelligence (AI) can support these collaborative efforts. Specifically, this thesis investigates how generative AI can enhance innovation processes and information sharing between government agencies and civil society organizations in projects addressing complex societal challenges. It examines the structural, cultural, and legal conditions that affect these collaborations and analyzes where and how AI tools might be implemented effectively. Based on semi-structured interviews with participants in the IBob-project, the study identifies four key areas that affect and hinders information sharing in cross-sector collaborations: differences in project initiation, differences in internal innovation processes, misaligned workflows, and legal uncertainty. While none of the participating organizations have implemented AI tools at the organizational level, all recognize the technology’s potential to improve and automate documentation, evaluation, and information sharing. They also emphasize the importance of a clear legal framework, cross-sector alignment, and both organizational and individual understanding of AI’s capabilities when implementing AI.