Making Society a Safer Place through Artificial Intelligence: The Slow-paced Transition from Idea to Sustainable Solution

While there has been vast research on how AI is used to automate and augment decision-making in organizational contexts, we engage a less developed, but equally important research agenda on how AI can address societal challenges. We present a public safety case of embracing AI and video surveillance to reduce incidents of drowning on harbor fronts in Denmark. We adopt Actor-Network Theory (ANT) and a qualitative inquiry to understand the complexities of this case. Our analysis reveals, how this AI-initiative, despite a shared goal of saving lives, faced setbacks and moved in a non-linear process from ideas to operation through networks of heterogeneous actors, humans as well as non-humans. By implication, our study sheds new light on the sociotechnical process of advancing AI initiatives for public safety and advocates for revitalizing ANT as a powerful framework in the age of AI.