Artificial intelligence and the judicial memory: the great misunderstanding

Abstract

Artificial intelligence is a widely debated topic, but it is often poorly defined. The purpose of this paper will be to return to the foundations of its functioning: mathematical and statistical formalism. While it works successfully in closed, quantifiable environments, it is far less adapted to open environments like the justice system. Indeed, the use of case law with these methods does not make it possible to enhance the judicial memory faithfully as the judgments that are processed do not necessarily include all of the grounds for a decision. This critique must not obscure another reality, which cuts across our whole society: an emerging project of governance by data, which is likely to call into question the very foundations of our democracies, including the rule of law.