Legal Judgment Reimagined: PredEx and the Rise of Intelligent AI Interpretation in Indian Courts

TL;DR

PredEx introduces a large, expert-annotated Indian CJPE dataset with over $15{,}222$ annotations to enable prediction with explanations. The study compares LM baselines and several instruction-tuned LLMs, showing that instruction tuning can improve explanatory depth while maintaining competitive judgment prediction performance. It demonstrates rigorous annotation quality control, cross-model evaluation (including expert ratings), and targeted prompting strategies, highlighting both the promise and current limitations of AI in explainable legal reasoning. The work sets a foundation for Indian-domain LLM development, stronger explainability, and future RLHF-based refinements to help reduce case backlog and improve transparency in the judiciary.

Abstract

In the era of Large Language Models (LLMs), predicting judicial outcomes poses significant challenges due to the complexity of legal proceedings and the scarcity of expert-annotated datasets. Addressing this, we introduce \textbf{Pred}iction with \textbf{Ex}planation (\texttt{PredEx}), the largest expert-annotated dataset for legal judgment prediction and explanation in the Indian context, featuring over 15,000 annotations. This groundbreaking corpus significantly enhances the training and evaluation of AI models in legal analysis, with innovations including the application of instruction tuning to LLMs. This method has markedly improved the predictive accuracy and explanatory depth of these models for legal judgments. We employed various transformer-based models, tailored for both general and Indian legal contexts. Through rigorous lexical, semantic, and expert assessments, our models effectively leverage \texttt{PredEx} to provide precise predictions and meaningful explanations, establishing it as a valuable benchmark for both the legal profession and the NLP community.

Figures (2)

• Figure 1: Illustration of the CJPE Task Framework.
• Figure 2: Annotated Example of Judicial Reasoning Extraction.