Increasing transparency around facial recognition technologies in law enforcement: towards a model framework

Abstract

Law enforcement authorities around the world are increasingly trialing or using facial recognition technologies (FRT). Their use has raised many legal and ethical challenges, one of which is a lack of transparency: Community members do not have sufficient information about what government organizations use FRT, for what purposes, and what safeguards are in place to manage the risks that they pose to human rights. While increased transparency around FRT use has been demanded by policy makers and academics, there are no established guidelines on how much transparency is needed around different FRT (authentication, identification, categorization) and what the barriers are in achieving the expected levels of transparency. This article fills this gap by proposing criteria which would help determine the required levels of transparency for different FRT applications (both low-risk and high-risk ones) and examines organizational, technical, legal and operational barriers in achieving adequate transparency and how they could be addressed.