Artificial intelligence and fair trial rights

Abstract

The right to a fair trial is the most frequently violated human right before international human rights bodies, and it is more the rule than the exception that national judicial systems are overburdened and overly slow. This chapter asks whether Artificial Intelligence(AI) and Machine Learning(ML) applications can help alleviate this problem, or if they are a threat to securing the right to the independent and impartial application of the law. It argues that the answer depends on whether the applications are designed with a clear vision of what courts are for and finds several current AI applications in various courts and public administrations to be missing this fundamental step. It identifies three key problems, namely the failure of current systems to differentiate between groups and individuals, the failure to take the fundamentally post factum nature of courts into account, and the tendency to abduct systems for another use than that which they were designed for. Following this the chapter builds a theoretical framework for determining what judge tasks can be allocated to or assisted by an AI application and which cannot. It argues that with careful application, cognitive computing type applications which extends the abilities of judges and clerks carries great potential in improving consistency and expediency of court cases. Finally, the chapter reviews emerging legislation on the usage of AI in judicial systems, and finds it to contain many of the same aims as the theoretical framework suggests incorporating, but to still lack detail for optimal application.