Keynote: Responsible AI in Policing

Responsible AI is deciding whether an AI tool should be used, as opposed to ‘could’ or ‘can’ it be used. And in deciding this, technical and statistical aspects of AI should not be separated from legal, contextual, operational, and ethical considerations. I will describe recent work on developing a responsible AI checklist for policing in the UK, touching on considerations of the broader judicial system. This is a collaboration between researchers in computer science, law, and social sciences.