Label Indeterminacy in AI & Law

TL;DR

This paper highlights the issue of label indeterminacy in legal machine learning, showing how unaccounted procedural variations can significantly influence model outcomes and emphasizing the need to address this challenge.

Contribution

It introduces the concept of label indeterminacy in AI & Law, demonstrating its impact on model behavior and discussing the limitations of existing imputation methods.

Findings

Label construction significantly affects model behavior.

Existing imputation methods rely on unverifiable assumptions.

Addressing label indeterminacy is crucial for legal ML applications.

Abstract

Machine learning is increasingly used in the legal domain, where it typically operates retrospectively by treating past case outcomes as ground truth. However, legal outcomes are often shaped by human interventions that are not captured in most machine learning approaches. A final decision may result from a settlement, an appeal, or other procedural actions. This creates label indeterminacy: the outcome could have been different if the intervention had or had not taken place. We argue that legal machine learning applications need to account for label indeterminacy. Methods exist that can impute these indeterminate labels, but they are all grounded in unverifiable assumptions. In the context of classifying cases from the European Court of Human Rights, we show that the way that labels are constructed during training can significantly affect model behaviour. We therefore position label indeterminacy as a relevant concern in AI & Law and demonstrate how it can shape model behaviour.