Legal AI

The project is being carried out as a collaboration between the Centre for Responsible AI, IIT Madras and Precog group, IIIT Hyderabad.

As AI models are increasingly becoming part of real world solutions across various sectors including the legal sector. AI-based Language Models for tasks like Judgement Prediction, have recently been proposed for the legal sector. Legal sector is a very crucial domain where errors can prove to be costly. Moreover, these models are strife with encoded social biases picked up from the training data. While bias and fairness have been studied across NLP, most studies primarily locate themselves within a Western context. Therefore, it is not only important to ensure that these models are deployed and monitored in a responsible manner, but also to ensure that the decisions/predictions generated are fair and ethical to all beneficiaries with a focus on Indian scenario as well. The project aims to explore techniques that can:

1. Detect fairness/bias issues in Legal AI models

2. Mitigate the bias observed in Legal AI models

Read our latest paper on arXiv here.