Nyaya Bot - AI Court Room Assistant for Justice Access
Abstract
References
Savelka and Ashley (Frontiers in AI, 2023) showed that LLMs like GPT-3.5 and GPT-4 can perform zero shot semantic annotation of legal texts, accurately identifying sentence roles without requiring fine-tuning.
Follow-up research by Savelka (arXiv, 2023) confirmed GPT’s effectiveness across contracts, statutes, and court opinions, demonstrating strong generalization in legal semantic labelling tasks.
A 2025 study in Symmetry proposed a Retrieval-Augmented Generation framework that improves the accuracy and reliability of legal text summarization through domain-specific grounding.
Together, these studies highlight the potential of modern LLMs to support practical legal workflows such as annotation, summarization, and intelligent legal assistance.
Advancing Access to Justice via Information and Communication Technology-World Bank report on ITC for closing the justice gap(Overview of tech-enabled justice reforms)
The accuracy, fairness, and limits of predicting recidivism - Dressel & Farid(2018) : empirical evaluation of COMPAS And human predictions.
Improving Access to Justice with Legal Chatbots - MDPI paper Describing chatbot prototypes and evaluation for legal help.
Machine Bias - ProPublica investigation into COMPAS and Algorithm bias in criminal justice.
Algorithm Fairness (Kleinberg et al.) - formal treatment of Fairness tradeoffs in criminal - justice algorithms.
EU AI ACT (overview/policy pages)-background and regulation classifying high-risk AI(implications for judicial use)
Refbacks
- There are currently no refbacks.