Explainable Artificial Intelligence for Assault Sentence Prediction in New Zealand

TL;DR

This paper explores the use of an explainable AI model to predict assault sentencing in New Zealand courts, demonstrating accuracy and interpretability, and discussing its potential benefits and risks.

Contribution

It introduces a proof-of-concept explainable AI model for assault sentence prediction and evaluates its accuracy and interpretability in a legal context.

Findings

Predicted sentences are accurate within one year.

The model identifies influential phrases affecting sentencing.

Discusses future benefits and risks of AI in courts.

Abstract

The judiciary has historically been conservative in its use of Artificial Intelligence, but recent advances in machine learning have prompted scholars to reconsider such use in tasks like sentence prediction. This paper investigates by experimentation the potential use of explainable artificial intelligence for predicting imprisonment sentences in assault cases in New Zealand's courts. We propose a proof-of-concept explainable model and verify in practice that it is fit for purpose, with predicted sentences accurate to within one year. We further analyse the model to understand the most influential phrases in sentence length prediction. We conclude the paper with an evaluative discussion of the future benefits and risks of different ways of using such an AI model in New Zealand's courts.