View on GitHub

fat2020tutorials.github.io

Can an algorithmic system be a ‘friend’ to a police officer’s discretion?

ACM FAT* 2020 Translation Tutorial

The overall goal of the tutorial is to increase understanding of the importance of discretion in police decision-making and to guide computer scientists in building algorithms that can be friends to that discretion. By the end of the tutorial, the audience will have explored:

  1. The legal concept of discretion in police decision-making in the context of common law, and potential legal challenges which could arise
  2. The realities and contexts in which police discretionary decisions are made, by reference to a MLA currently in development within a domestic violence context, and the views of victims, offenders and police officers
  3. Algorithmic design decisions/features which may help or hinder the legitimate exercise of police discretion
  4. Implications for the choice of data inputs, explanation of the output and the means by which it was obtained, user interface design and the interaction of the algorithmic system with other non-codified relevant information.

Who will be delivering the tutorial?

Who is the target audience?

Anyone interested in considering practical, legal and technical issues crucial to avoiding the prejudicial and instead developing algorithms that are supportive – a ‘friend’– to legitimate police and other public sector discretionary decision-making.

Do I need to prepare anything before attending?

None. Just a willingness to participate and get involved in discussion!

Slides

Slides available here.