ukresrisk
Abstract
“Good old fashioned” AI, developed first in the 1980s but still an approach used in many contemporary legal apps and law chatbots, is often seen as less likely to create a dangerous “black box society” than machine learning based approaches. The chapter queries this notion by looking at the way in which the very process of formalising the law rests on normative decisions and value commitments that can’t simply be left to software developers. Using the literary figure of the Golem, it traces some of the normative decisions that any legal technology has to make, and posits some desiderata for an ethically responsible theory of legal formalisation.
